# Lessons from Open Python2 LDSC Issues (Bulik/ldsc)
As of March 01, 2026, the upstream Python2 repository has **318 open issues**. This document distills recurring pain points into actionable lessons for the Rust rewrite, and calls out Python2-specific problems that the Rust design avoids.
**Note:** Issue categorization below is heuristic (title/body keyword-based) to ensure all open issues are covered. Manual review is recommended before prioritizing changes.
## Cross-Cutting Lessons (Themes)
- **Regression Output/Interpretation**: 117 open issues. Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable.
- **Input Format/Parsing**: 83 open issues. Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes.
- **Annotations/Partitioned**: 33 open issues. Clarify annotation formats, baseline model expectations, and overlap rules; add validation that the annotation/LD score matrices align.
- **Reference Data/Downloads**: 23 open issues. Document stable reference file sources, versions, and expected contents; add preflight checks and mirror guidance for common downloads.
- **LD Score Computation/Performance**: 22 open issues. Expose memory/perf guidance and progress; ensure robust error handling for large datasets and edge-case inputs.
- **Install/Env**: 16 open issues. Provide zero-setup binaries/containers and precise dependency checks; surface clear diagnostics for missing system libs and avoid Python/conda/pandas issues.
- **Other/Uncategorized**: 11 open issues. Document the requested workflow or clarify constraints with a concrete example.
- **Feature/UX**: 10 open issues. Add guidance or tooling for requested workflows (plots, tutorials, web resources) and clearly state what is out of scope.
- **Population/Genome Build**: 3 open issues. Call out reference genome builds and population-specific resources; warn when mixing builds or using unsupported populations.
## Recommended Documentation/UX Additions (for New Users)
- **One-page quickstart**: minimal example for `munge-sumstats`, `l2`, `h2`, and `rg`, with file naming conventions and expected outputs.
- **Input format FAQ**: required columns, acceptable synonyms, gzip detection, and common parsing errors with examples.
- **Reference data matrix**: which reference panels to use by population and genome build, with verified download URLs and checksums.
- **Troubleshooting guide**: NA/negative/rg>1 outcomes, low h2, and intercept interpretation with decision trees.
- **Annotation/CTS walkthrough**: end-to-end example for baseline + custom annotations, including expected file shapes.
- **Performance guidance**: memory expectations by dataset size, and thread tuning flags for LD score computation.
## Problems in Python2 LDSC That the Rust Rewrite Avoids by Design
- **Python 2/3 environment fragility**: conda/pip install failures, pandas version constraints, and `bz2` module issues go away with a single compiled binary or container.
- **Pandas-specific indexing/dtype errors**: common `KeyError`, `TypeError`, and `FutureWarning` messages from pandas/numpy are avoided by Rust-native parsing.
- **Interpreter/runtime warnings**: numpy `rcond` warnings and Python runtime warnings are not applicable in Rust.
- **Tab/indentation and script mismatch issues**: Python source-level indentation errors and mixed-script version drift are avoided with a single Rust CLI.
- **Opaque gzip/text decoding errors**: Rust can detect compression by magic bytes and produce explicit diagnostics rather than Python’s generic parse errors.
## Issue-by-Issue Lessons (All Open Issues)
Each row links the upstream issue and states the most relevant lesson for the Rust rewrite documentation/UX.
| Issue | Title | Theme | Lesson for Rust Rewrite |
| --- | --- | --- | --- |
| [#496](https://github.com/bulik/ldsc/issues/496) | LDSC Genetic Correlation Analysis Returns All NA Values - Seeking Help | Regression Output/Interpretation | Add diagnostics for NA outputs (e.g., negative jackknife blocks, low h2, or missing overlap). |
| [#495](https://github.com/bulik/ldsc/issues/495) | Quality Control when converting 1000 Genome Project Phase 3 vcf data into plink file | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#494](https://github.com/bulik/ldsc/issues/494) | negative MAF/LD Score Correlation | LD Score Computation/Performance | Explain expected MAF/LD correlation ranges and note that small negatives can occur due to noise. |
| [#492](https://github.com/bulik/ldsc/issues/492) | Segmentation fault | LD Score Computation/Performance | Harden against crashes and emit a clear, actionable error message when internal computation fails. |
| [#491](https://github.com/bulik/ldsc/issues/491) | Number of variants included for computing LD score at each SNP | LD Score Computation/Performance | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#490](https://github.com/bulik/ldsc/issues/490) | Error running munge_sumstats.py | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#489](https://github.com/bulik/ldsc/issues/489) | Is it possible to run heritability and/or partitioned heritability with just p values and no signed-sumstats? And if yes - how? | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#488](https://github.com/bulik/ldsc/issues/488) | SNPs included in the pre-computed LD scores for European are much less than SNPs in 1000 genome phase 1 dataset | LD Score Computation/Performance | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#484](https://github.com/bulik/ldsc/issues/484) | Clarification on Baseline Model and Obtaining Enrichment Scores for Cell-Type-Specific Heritability (e.g., with scATAC-seq annotations) | Regression Output/Interpretation | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#483](https://github.com/bulik/ldsc/issues/483) | Error with calculating LD score using w_hm3.snplist: After merging with --print-snps, no SNPs remain. | Input Format/Parsing | Clarify `--print-snps` input format (single SNP column) and show expected overlap behavior. |
| [#481](https://github.com/bulik/ldsc/issues/481) | what is the value in parenthesis in Total Observed scale h2? | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#479](https://github.com/bulik/ldsc/issues/479) | How to make customize `panel` ? | Annotations/Partitioned | Clarify annotation formats, baseline model expectations, and overlap rules; add validation that the annotation/LD score matrices align. |
| [#478](https://github.com/bulik/ldsc/issues/478) | ldsc python 3 H2 estimation question | Regression Output/Interpretation | Emphasize the Rust binary removes Python/conda dependency issues; document minimal system requirements. |
| [#476](https://github.com/bulik/ldsc/issues/476) | Baseline LD scores | Reference Data/Downloads | Explain baseline vs full baselineLD usage and provide a default recommended model. |
| [#475](https://github.com/bulik/ldsc/issues/475) | LDSC negative heritability | Regression Output/Interpretation | Explain negative estimates and when to treat h2/rg as unreliable; suggest checking sample size and QC. |
| [#472](https://github.com/bulik/ldsc/issues/472) | rg values not in the [-1,1] bound&NAN Values | Regression Output/Interpretation | Explain why rg can exceed [-1,1] due to noise and provide interpretation guidance. |
| [#471](https://github.com/bulik/ldsc/issues/471) | how to build my own input weight files for --w-ld-chr | Input Format/Parsing | Ensure required columns like `CHR` are present or allow column mapping; error should list required columns and example headers. |
| [#470](https://github.com/bulik/ldsc/issues/470) | LDSC on build38 data | Population/Genome Build | Document GRCh38 support status and provide compatible reference panels or conversion guidance. |
| [#469](https://github.com/bulik/ldsc/issues/469) | LDSC for each of the 1703 regions defined by ldetech (Joseph K Pickrell 2016) | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#467](https://github.com/bulik/ldsc/issues/467) | Does immunochip summary data with another gwas summary data can be used to LDSC genetic correlation? | Regression Output/Interpretation | Clarify scope limitations for non-standard variant types or species, and suggest alternatives. |
| [#466](https://github.com/bulik/ldsc/issues/466) | Index error/ list index out of range | LD Score Computation/Performance | Expose memory/perf guidance and progress; ensure robust error handling for large datasets and edge-case inputs. |
| [#465](https://github.com/bulik/ldsc/issues/465) | LD Hub Website Inaccessibility | Feature/UX | Add guidance or tooling for requested workflows (plots, tutorials, web resources) and clearly state what is out of scope. |
| [#464](https://github.com/bulik/ldsc/issues/464) | Heritability Estimation Error!! | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#463](https://github.com/bulik/ldsc/issues/463) | How to calculate stratified-LDSC between two traits in special function groups? | Other/Uncategorized | Document the requested workflow or clarify constraints with a concrete example. |
| [#462](https://github.com/bulik/ldsc/issues/462) | How to generate one ENSG_coord.txt file for the make_annot.py ? | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#461](https://github.com/bulik/ldsc/issues/461) | How to generate one own cts annotation file? | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#460](https://github.com/bulik/ldsc/issues/460) | customized expression data for cell specific analysis | Annotations/Partitioned | Clarify annotation formats, baseline model expectations, and overlap rules; add validation that the annotation/LD score matrices align. |
| [#458](https://github.com/bulik/ldsc/issues/458) | Genetic correlation > 1 | Regression Output/Interpretation | Explain why rg can exceed [-1,1] due to noise and provide interpretation guidance. |
| [#457](https://github.com/bulik/ldsc/issues/457) | Extension of S-LDSC | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#456](https://github.com/bulik/ldsc/issues/456) | valid alleles QC | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#454](https://github.com/bulik/ldsc/issues/454) | Other species | Population/Genome Build | Clarify scope limitations for non-standard variant types or species, and suggest alternatives. |
| [#452](https://github.com/bulik/ldsc/issues/452) | Is enrichment same when using LD scores with different combination of categories? | Annotations/Partitioned | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#450](https://github.com/bulik/ldsc/issues/450) | Cannot run tutorial | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#447](https://github.com/bulik/ldsc/issues/447) | Questions Regarding S-LDSC Analysis with --ref-ld-chr-cts | Input Format/Parsing | Ensure required columns like `CHR` are present or allow column mapping; error should list required columns and example headers. |
| [#446](https://github.com/bulik/ldsc/issues/446) | munge_sumstats issue: Usecols do not match names | Input Format/Parsing | Detect empty or malformed files early and print the expected header and delimiter examples. |
| [#444](https://github.com/bulik/ldsc/issues/444) | Estimating bivariate LDSC intercept when h2 is low | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#443](https://github.com/bulik/ldsc/issues/443) | Error # terms in --M must match # of LD Scores in --ref-ld | LD Score Computation/Performance | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#442](https://github.com/bulik/ldsc/issues/442) | Syntax Error | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#439](https://github.com/bulik/ldsc/issues/439) | Filtering of munge_sumstats.py | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#438](https://github.com/bulik/ldsc/issues/438) | Prop._h2 is negative | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#437](https://github.com/bulik/ldsc/issues/437) | Partition heritability for UKB data | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#436](https://github.com/bulik/ldsc/issues/436) | Filtration of partition heritability results | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#435](https://github.com/bulik/ldsc/issues/435) | Error parsing .annot file | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#434](https://github.com/bulik/ldsc/issues/434) | Lowest limit of rg p-value reported? | Regression Output/Interpretation | Handle extreme or non-numeric p-values (very small exponents, OR columns); provide guidance to clamp or convert. |
| [#433](https://github.com/bulik/ldsc/issues/433) | African data analysis | Population/Genome Build | Call out reference genome builds and population-specific resources; warn when mixing builds or using unsupported populations. |
| [#432](https://github.com/bulik/ldsc/issues/432) | Is python 2.7 no longer available to download using code in ldsc tutorial? | Reference Data/Downloads | Maintain updated download mirrors and checksums for reference data and tutorial files. |
| [#431](https://github.com/bulik/ldsc/issues/431) | INTERCEPT FLAG | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#430](https://github.com/bulik/ldsc/issues/430) | p values in version Version 1.0.1 are different from v 2.0.0 | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#429](https://github.com/bulik/ldsc/issues/429) | https://data.broadinstitute.org/alkesgroup/LDSCORE/eur_w_ld_chr.tar.bz2 download problem? | Reference Data/Downloads | Maintain updated download mirrors and checksums for reference data and tutorial files. |
| [#428](https://github.com/bulik/ldsc/issues/428) | where can i find ENSG_coord.txt or similiar plain file in different build? | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#426](https://github.com/bulik/ldsc/issues/426) | MemoryError: When running LDSCORE analysis | LD Score Computation/Performance | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#425](https://github.com/bulik/ldsc/issues/425) | Is it necessary to use the --merge-alleles parameter when doing genetic correlation in ldsc? | Regression Output/Interpretation | Document `--merge-alleles` format (SNP, A1, A2) and show how allele mismatches are handled. |
| [#424](https://github.com/bulik/ldsc/issues/424) | Missing frequency from GWAS | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#423](https://github.com/bulik/ldsc/issues/423) | ./ldsc.py -h Illegal instruction (core dumped) | LD Score Computation/Performance | Document minimum CPU feature requirements and provide a compatible build if needed. |
| [#422](https://github.com/bulik/ldsc/issues/422) | NotImplementedError: iLocation based boolean indexing on an integer type is not available | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#421](https://github.com/bulik/ldsc/issues/421) | lsdc :IndexError: list index out of range | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#420](https://github.com/bulik/ldsc/issues/420) | munge_sumstats.py taking forever to complete analysis | Input Format/Parsing | Provide runtime expectations and guidance on thread tuning and I/O bottlenecks. |
| [#419](https://github.com/bulik/ldsc/issues/419) | rg with sample overlap | Regression Output/Interpretation | Explain how to handle sample overlap and which flags/inputs are required. |
| [#418](https://github.com/bulik/ldsc/issues/418) | ERROR: P<1e-300 | Input Format/Parsing | Handle extreme or non-numeric p-values (very small exponents, OR columns); provide guidance to clamp or convert. |
| [#417](https://github.com/bulik/ldsc/issues/417) | Can't run ldsc under screen at Linux | Install/Env | Provide zero-setup binaries/containers and precise dependency checks; surface clear diagnostics for missing system libs and avoid Python/conda/pandas issues. |
| [#416](https://github.com/bulik/ldsc/issues/416) | how to get the filtered snps | LD Score Computation/Performance | Expose memory/perf guidance and progress; ensure robust error handling for large datasets and edge-case inputs. |
| [#415](https://github.com/bulik/ldsc/issues/415) | Sample size with repeated observations | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#414](https://github.com/bulik/ldsc/issues/414) | Need LDSC file | Reference Data/Downloads | Document stable reference file sources, versions, and expected contents; add preflight checks and mirror guidance for common downloads. |
| [#413](https://github.com/bulik/ldsc/issues/413) | How to draw the visialization of the LDSC regression plot like its NG paper? | Feature/UX | Add guidance or tooling for requested workflows (plots, tutorials, web resources) and clearly state what is out of scope. |
| [#412](https://github.com/bulik/ldsc/issues/412) | Cell type specific analyses | Reference Data/Downloads | Document stable reference file sources, versions, and expected contents; add preflight checks and mirror guidance for common downloads. |
| [#411](https://github.com/bulik/ldsc/issues/411) | Partitioned genetic correlation | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#410](https://github.com/bulik/ldsc/issues/410) | correlation value higher than 1 | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#409](https://github.com/bulik/ldsc/issues/409) | bucket seems to be unavailable | Reference Data/Downloads | Maintain updated download mirrors and checksums for reference data and tutorial files. |
| [#408](https://github.com/bulik/ldsc/issues/408) | Error running LD score regression | LD Score Computation/Performance | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#407](https://github.com/bulik/ldsc/issues/407) | Suitable w_hm3.snplist for AFR? | Reference Data/Downloads | Document HapMap3 SNP list usage and where to obtain the correct version for each build. |
| [#406](https://github.com/bulik/ldsc/issues/406) | Add CM and MAF file to LD Score output .gz | Input Format/Parsing | Explain expected MAF/LD correlation ranges and note that small negatives can occur due to noise. |
| [#405](https://github.com/bulik/ldsc/issues/405) | 'Could not find a signed summary statistic column.') | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#403](https://github.com/bulik/ldsc/issues/403) | 1kg_eur.tar.bz2 not found | Reference Data/Downloads | Maintain updated download mirrors and checksums for reference data and tutorial files. |
| [#402](https://github.com/bulik/ldsc/issues/402) | Could anyone send me the "eas_ldscores.tar.bz2"? I wanted to calculate the ldsc of Asia, but I found I didn't have such file. And I can't download it from cloud. If anyone can send me it, my e-mail is chenqiwjc@gmail.com. Thank you so much. | Reference Data/Downloads | Maintain updated download mirrors and checksums for reference data and tutorial files. |
| [#401](https://github.com/bulik/ldsc/issues/401) | how to add annotation to the baseline model??? | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#400](https://github.com/bulik/ldsc/issues/400) | TypeError: a bytes-like object is required, not 'str' | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#396](https://github.com/bulik/ldsc/issues/396) | --overlap-annot does not recognize path in --annot | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#395](https://github.com/bulik/ldsc/issues/395) | tutorial request | Feature/UX | Add guidance or tooling for requested workflows (plots, tutorials, web resources) and clearly state what is out of scope. |
| [#393](https://github.com/bulik/ldsc/issues/393) | Why is the number of SNPs in a cell type the same to the control's one? | Annotations/Partitioned | Clarify annotation formats, baseline model expectations, and overlap rules; add validation that the annotation/LD score matrices align. |
| [#390](https://github.com/bulik/ldsc/issues/390) | How to test the null hypothesis of rg=1 | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#389](https://github.com/bulik/ldsc/issues/389) | Error calculating heritability | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#388](https://github.com/bulik/ldsc/issues/388) | Question about partitioned LDSC | Annotations/Partitioned | Clarify annotation formats, baseline model expectations, and overlap rules; add validation that the annotation/LD score matrices align. |
| [#387](https://github.com/bulik/ldsc/issues/387) | merge-alleles: command not found | Input Format/Parsing | Maintain updated download mirrors and checksums for reference data and tutorial files. |
| [#386](https://github.com/bulik/ldsc/issues/386) | Where to obtain "hapmap3_snps.tgz"? | Reference Data/Downloads | Document HapMap3 SNP list usage and where to obtain the correct version for each build. |
| [#385](https://github.com/bulik/ldsc/issues/385) | Error: Missing some LD scores from cts files. Are you sure all SNPs in ref-ld-chr are also in ref-ld-chr-cts | Input Format/Parsing | Ensure required columns like `CHR` are present or allow column mapping; error should list required columns and example headers. |
| [#384](https://github.com/bulik/ldsc/issues/384) | How to obtain the file "w_hm3.snplist.bz2" | Reference Data/Downloads | Document HapMap3 SNP list usage and where to obtain the correct version for each build. |
| [#383](https://github.com/bulik/ldsc/issues/383) | Could I set the number of chromosomes in LDSC? | Annotations/Partitioned | Clarify annotation formats, baseline model expectations, and overlap rules; add validation that the annotation/LD score matrices align. |
| [#380](https://github.com/bulik/ldsc/issues/380) | LDSC requires pandas version >= 0.17.0 | Install/Env | Emphasize the Rust binary removes Python/conda dependency issues; document minimal system requirements. |
| [#379](https://github.com/bulik/ldsc/issues/379) | LDSC requires pandas version >= 0.17.0 | Install/Env | Emphasize the Rust binary removes Python/conda dependency issues; document minimal system requirements. |
| [#378](https://github.com/bulik/ldsc/issues/378) | Unable to download from google bucket | Reference Data/Downloads | Maintain updated download mirrors and checksums for reference data and tutorial files. |
| [#377](https://github.com/bulik/ldsc/issues/377) | Reference files - 1000G LD scores | Reference Data/Downloads | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#376](https://github.com/bulik/ldsc/issues/376) | weights for non-European populations | Reference Data/Downloads | Explain the expected weight LD score files and how to build custom weights. |
| [#375](https://github.com/bulik/ldsc/issues/375) | About A1 / A2 allele definition in wiki and `munge_sumstats.py` | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#374](https://github.com/bulik/ldsc/issues/374) | Why do I keep reporting an error: Could not find SNP column？ | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#373](https://github.com/bulik/ldsc/issues/373) | Presence of ldsc 2.0.1 on PyPI but not on GitHub | Install/Env | Provide zero-setup binaries/containers and precise dependency checks; surface clear diagnostics for missing system libs and avoid Python/conda/pandas issues. |
| [#372](https://github.com/bulik/ldsc/issues/372) | No module named bz2 | Install/Env | Emphasize the Rust binary removes Python/conda dependency issues; document minimal system requirements. |
| [#371](https://github.com/bulik/ldsc/issues/371) | link to reference files does not work | Reference Data/Downloads | Maintain updated download mirrors and checksums for reference data and tutorial files. |
| [#370](https://github.com/bulik/ldsc/issues/370) | questions about pre-computed ldscore | LD Score Computation/Performance | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#369](https://github.com/bulik/ldsc/issues/369) | Getting module parse has no attibute error | Input Format/Parsing | Emphasize the Rust binary removes Python/conda dependency issues; document minimal system requirements. |
| [#367](https://github.com/bulik/ldsc/issues/367) | parse.py error? | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#366](https://github.com/bulik/ldsc/issues/366) | IndexError: list index out of range | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#365](https://github.com/bulik/ldsc/issues/365) | Unalignable boolean Series provided as indexer (index of the boolean Series and of the indexed object do not match). | Install/Env | Provide zero-setup binaries/containers and precise dependency checks; surface clear diagnostics for missing system libs and avoid Python/conda/pandas issues. |
| [#364](https://github.com/bulik/ldsc/issues/364) | how to make my own baseline model? | Annotations/Partitioned | Explain baseline vs full baselineLD usage and provide a default recommended model. |
| [#363](https://github.com/bulik/ldsc/issues/363) | Sample overlap | Regression Output/Interpretation | Explain how to handle sample overlap and which flags/inputs are required. |
| [#362](https://github.com/bulik/ldsc/issues/362) | h2 estimate and intercept | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#361](https://github.com/bulik/ldsc/issues/361) | make_annot.py_error "AttributeError: 'float' object has no attribute 'isdigit'" | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#359](https://github.com/bulik/ldsc/issues/359) | LDSC is not available for Python 3 | Install/Env | Emphasize the Rust binary removes Python/conda dependency issues; document minimal system requirements. |
| [#358](https://github.com/bulik/ldsc/issues/358) | munge_sumstats.py extremely slow runtime in the provided conda environment | Input Format/Parsing | Provide runtime expectations and guidance on thread tuning and I/O bottlenecks. |
| [#357](https://github.com/bulik/ldsc/issues/357) | w_hm3.snplist for GRCh38 | Reference Data/Downloads | Document GRCh38 support status and provide compatible reference panels or conversion guidance. |
| [#356](https://github.com/bulik/ldsc/issues/356) | Error converting summary statistics | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#355](https://github.com/bulik/ldsc/issues/355) | UnboundLocalError: local variable 'cname_translation' referenced before assignment | Reference Data/Downloads | Document stable reference file sources, versions, and expected contents; add preflight checks and mirror guidance for common downloads. |
| [#354](https://github.com/bulik/ldsc/issues/354) | TypeError: can only concatenate list (not "dict_keys") to list | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#353](https://github.com/bulik/ldsc/issues/353) | After merging with reference panel LD, 0 SNPs remain | Annotations/Partitioned | Clarify annotation formats, baseline model expectations, and overlap rules; add validation that the annotation/LD score matrices align. |
| [#352](https://github.com/bulik/ldsc/issues/352) | ValueError: LD Scores for concatenation must have identical SNP columns. | Input Format/Parsing | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#351](https://github.com/bulik/ldsc/issues/351) | TabError: inconsistent use of tabs and spaces in indentation | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#350](https://github.com/bulik/ldsc/issues/350) | Pybedtools not available for windows os | Install/Env | Provide zero-setup binaries/containers and precise dependency checks; surface clear diagnostics for missing system libs and avoid Python/conda/pandas issues. |
| [#349](https://github.com/bulik/ldsc/issues/349) | Cell type specific analyses index error | LD Score Computation/Performance | Expose memory/perf guidance and progress; ensure robust error handling for large datasets and edge-case inputs. |
| [#348](https://github.com/bulik/ldsc/issues/348) | Getting Nan in the MAF/LD Score Correlation Matrix | Regression Output/Interpretation | Explain expected MAF/LD correlation ranges and note that small negatives can occur due to noise. |
| [#346](https://github.com/bulik/ldsc/issues/346) | Sorry wrong github | Other/Uncategorized | Document the requested workflow or clarify constraints with a concrete example. |
| [#345](https://github.com/bulik/ldsc/issues/345) | FloatingPointError: invalid value encountered in divide | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#342](https://github.com/bulik/ldsc/issues/342) | Bug: l2_parser can fail to load ldscores | Input Format/Parsing | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#340](https://github.com/bulik/ldsc/issues/340) | Using LDSC for CNV-GWAS | Regression Output/Interpretation | Clarify scope limitations for non-standard variant types or species, and suggest alternatives. |
| [#339](https://github.com/bulik/ldsc/issues/339) | Variant filtering (not SNPs or strand-ambiguous) | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#337](https://github.com/bulik/ldsc/issues/337) | Can't find EAS HapMap3 SNPs in alkesgroup repository | Reference Data/Downloads | Document HapMap3 SNP list usage and where to obtain the correct version for each build. |
| [#336](https://github.com/bulik/ldsc/issues/336) | tau* z-scores | Other/Uncategorized | Document the requested workflow or clarify constraints with a concrete example. |
| [#335](https://github.com/bulik/ldsc/issues/335) | `--n-min` default in args description does not reflect underlying code | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#334](https://github.com/bulik/ldsc/issues/334) | BOLT-LMM v.2.3.5 in combination with LDSC | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#333](https://github.com/bulik/ldsc/issues/333) | problems on --merge-alleles | Input Format/Parsing | Document `--merge-alleles` format (SNP, A1, A2) and show how allele mismatches are handled. |
| [#332](https://github.com/bulik/ldsc/issues/332) | Interpretation of genetic correlation between a continuous trait and a binary trait | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#331](https://github.com/bulik/ldsc/issues/331) | 1000G_Phase3_plinkfiles.tgz - GRCh37 or GRCh38 | Input Format/Parsing | Document GRCh38 support status and provide compatible reference panels or conversion guidance. |
| [#330](https://github.com/bulik/ldsc/issues/330) | HTTP request sent, awaiting response... 404 Not Found | Feature/UX | Maintain updated download mirrors and checksums for reference data and tutorial files. |
| [#329](https://github.com/bulik/ldsc/issues/329) | Munge Error - ValueError: could not convert string to float: OR | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#328](https://github.com/bulik/ldsc/issues/328) | High genetic correlation, but large se and not significant, h2 is weirdly low too | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#327](https://github.com/bulik/ldsc/issues/327) | munge_sumstats.py_error "ValueError: No object to concatenate" | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#325](https://github.com/bulik/ldsc/issues/325) | LD Hub | Feature/UX | Add guidance or tooling for requested workflows (plots, tutorials, web resources) and clearly state what is out of scope. |
| [#324](https://github.com/bulik/ldsc/issues/324) | Problems when perform LD score regression by --rg | Regression Output/Interpretation | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#323](https://github.com/bulik/ldsc/issues/323) | Using stratified model for Rg (old_weight is always FALSE for --rg) | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#322](https://github.com/bulik/ldsc/issues/322) | HapMap3 variants | Reference Data/Downloads | Document HapMap3 SNP list usage and where to obtain the correct version for each build. |
| [#321](https://github.com/bulik/ldsc/issues/321) | How to replicate pre-computed LD score | LD Score Computation/Performance | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#320](https://github.com/bulik/ldsc/issues/320) | Which --signed-sumstats? And NA genetic correlation | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#318](https://github.com/bulik/ldsc/issues/318) | Partitioned heritability: MAF_Adj_LLD_AFR annotation | Regression Output/Interpretation | Explain expected MAF/LD correlation ranges and note that small negatives can occur due to noise. |
| [#317](https://github.com/bulik/ldsc/issues/317) | rg –intercept-h2: constrain the intercept of a single sumstat | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#316](https://github.com/bulik/ldsc/issues/316) | IOError: CRC check failed 0xf157293c != 0xb6a8ec8fL | Input Format/Parsing | Auto-detect compression and warn when file extensions do not match actual compression. |
| [#315](https://github.com/bulik/ldsc/issues/315) | IndexError: list index out of range \| ValueError: Found 2 different SNP column | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#314](https://github.com/bulik/ldsc/issues/314) | Partitioned heritability - comparing enrichment between two categories | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#313](https://github.com/bulik/ldsc/issues/313) | High ratio, and high h2 confidence interval in comparison with UKBB | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#311](https://github.com/bulik/ldsc/issues/311) | LD Hub website issues | Feature/UX | Add guidance or tooling for requested workflows (plots, tutorials, web resources) and clearly state what is out of scope. |
| [#310](https://github.com/bulik/ldsc/issues/310) | Can I apply ldsc to small sample size (N<1000) for genetic correlation? | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#309](https://github.com/bulik/ldsc/issues/309) | error | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#308](https://github.com/bulik/ldsc/issues/308) | Very high Ratio, negative h2, and doubt about constraining the intercept | Regression Output/Interpretation | Explain negative estimates and when to treat h2/rg as unreliable; suggest checking sample size and QC. |
| [#305](https://github.com/bulik/ldsc/issues/305) | How should I run LDSC on a small annotation? | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#304](https://github.com/bulik/ldsc/issues/304) | --N flag ignored by munge_sumstats.py | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#303](https://github.com/bulik/ldsc/issues/303) | specifying columns in annotation file | Input Format/Parsing | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#302](https://github.com/bulik/ldsc/issues/302) | LDSC for structural variants | Install/Env | Clarify scope limitations for non-standard variant types or species, and suggest alternatives. |
| [#301](https://github.com/bulik/ldsc/issues/301) | nothing | Other/Uncategorized | Document the requested workflow or clarify constraints with a concrete example. |
| [#300](https://github.com/bulik/ldsc/issues/300) | Problem with irwls | LD Score Computation/Performance | Expose memory/perf guidance and progress; ensure robust error handling for large datasets and edge-case inputs. |
| [#299](https://github.com/bulik/ldsc/issues/299) | LDSC with QTL summary statistics | Other/Uncategorized | Document the requested workflow or clarify constraints with a concrete example. |
| [#298](https://github.com/bulik/ldsc/issues/298) | rg all vs all | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#297](https://github.com/bulik/ldsc/issues/297) | How to use own gene set obtained from single cell data to do cell type specific analysis? | Other/Uncategorized | Document the requested workflow or clarify constraints with a concrete example. |
| [#296](https://github.com/bulik/ldsc/issues/296) | munge_sumstats in python3 | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#295](https://github.com/bulik/ldsc/issues/295) | Can I calculate ld scores from my own reference samples? | Reference Data/Downloads | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#293](https://github.com/bulik/ldsc/issues/293) | LD score regression of data including SNPs with extremely low P-value | Input Format/Parsing | Handle extreme or non-numeric p-values (very small exponents, OR columns); provide guidance to clamp or convert. |
| [#292](https://github.com/bulik/ldsc/issues/292) | munge_sumstats.py takes forever with file contains more than 1M lines. | Input Format/Parsing | Provide runtime expectations and guidance on thread tuning and I/O bottlenecks. |
| [#290](https://github.com/bulik/ldsc/issues/290) | Some results have NA values | Regression Output/Interpretation | Add diagnostics for NA outputs (e.g., negative jackknife blocks, low h2, or missing overlap). |
| [#288](https://github.com/bulik/ldsc/issues/288) | One-sided test with --h2 flag and custom annotations | Regression Output/Interpretation | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#287](https://github.com/bulik/ldsc/issues/287) | ERROR: Command errored out with exit status 1: /Users/hongyan/opt/anaconda3/envs/ldsc/bin/python -u -c 'import sys, setuptools, tokenize; sys.argv[0] = '"'"'/private/var/folders/9t/965z56gn1lj_tknr10tt47w80000gn/T/pip-install-3eZFqg/pandas/setup.py'"'"'; __file__='"'"'/private/var/folders/9t/965z56gn1lj_tknr10tt47w80000gn/T/pip-install-3eZFqg/pandas/setup.py'"'"';f=getattr(tokenize, '"'"'open'"'"', open)(__file__);code=f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, __file__, '"'"'exec'"'"'))' install --record /private/var/folders/9t/965z56gn1lj_tknr10tt47w80000gn/T/pip-record-ltQTy5/install-record.txt --single-version-externally-managed --compile --install-headers /Users/hongyan/opt/anaconda3/envs/ldsc/include/python2.7/pandas Check the logs for full command output. failed CondaEnvException: Pip failed | Install/Env | Emphasize the Rust binary removes Python/conda dependency issues; document minimal system requirements. |
| [#286](https://github.com/bulik/ldsc/issues/286) | Minimal set of rsIDs for heritability estimation (not an issue) | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#285](https://github.com/bulik/ldsc/issues/285) | Different h2 results for the same phenotype | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#284](https://github.com/bulik/ldsc/issues/284) | genetic enrichment of 3'UTR | Annotations/Partitioned | Clarify annotation formats, baseline model expectations, and overlap rules; add validation that the annotation/LD score matrices align. |
| [#283](https://github.com/bulik/ldsc/issues/283) | munge_sumstats.py error: questions for input file | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#282](https://github.com/bulik/ldsc/issues/282) | Data for tutorial no longer available | Feature/UX | Add guidance or tooling for requested workflows (plots, tutorials, web resources) and clearly state what is out of scope. |
| [#281](https://github.com/bulik/ldsc/issues/281) | mungestats takes a long time | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#280](https://github.com/bulik/ldsc/issues/280) | Little variations in intercept value | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#278](https://github.com/bulik/ldsc/issues/278) | annot.gz file not generated when running LD Score Estimation | Input Format/Parsing | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#277](https://github.com/bulik/ldsc/issues/277) | CondaEnvException: Pip failed | Install/Env | Emphasize the Rust binary removes Python/conda dependency issues; document minimal system requirements. |
| [#275](https://github.com/bulik/ldsc/issues/275) | munge_sumstats error on encountering low p-values. Implications for genetic correlation? | Regression Output/Interpretation | Handle extreme or non-numeric p-values (very small exponents, OR columns); provide guidance to clamp or convert. |
| [#274](https://github.com/bulik/ldsc/issues/274) | ValueError: Missing some LD scores from cts files. Are you sure all SNPs in ref-ld-chr are also in ref-ld-chr-cts | Input Format/Parsing | Ensure required columns like `CHR` are present or allow column mapping; error should list required columns and example headers. |
| [#273](https://github.com/bulik/ldsc/issues/273) | Heritability tutorial download links broken | Regression Output/Interpretation | Maintain updated download mirrors and checksums for reference data and tutorial files. |
| [#272](https://github.com/bulik/ldsc/issues/272) | munge_sumstats.py column names specified on the command line should override any defaults | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#271](https://github.com/bulik/ldsc/issues/271) | interpreting --a1-inc | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#269](https://github.com/bulik/ldsc/issues/269) | IOError: Not a gzipped file | Input Format/Parsing | Auto-detect compression and warn when file extensions do not match actual compression. |
| [#268](https://github.com/bulik/ldsc/issues/268) | It took long time to run munge_sumstats.py -- | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#265](https://github.com/bulik/ldsc/issues/265) | Low polygenicity signal | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#264](https://github.com/bulik/ldsc/issues/264) | What if GWAS summary statistics are from a different reference genome build? | Reference Data/Downloads | Document stable reference file sources, versions, and expected contents; add preflight checks and mirror guidance for common downloads. |
| [#262](https://github.com/bulik/ldsc/issues/262) | RG>1 && sumsates.gz without column P | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#259](https://github.com/bulik/ldsc/issues/259) | Rg >1? Rg=0.84 but P non-significant | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#258](https://github.com/bulik/ldsc/issues/258) | MemoryError | LD Score Computation/Performance | Publish memory requirements and provide tips to reduce memory footprint (chromosome split, smaller windows). |
| [#257](https://github.com/bulik/ldsc/issues/257) | 2 architecture questions on LDSC | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#255](https://github.com/bulik/ldsc/issues/255) | rg was out of bounds. | Regression Output/Interpretation | Explain why rg can exceed [-1,1] due to noise and provide interpretation guidance. |
| [#254](https://github.com/bulik/ldsc/issues/254) | No results file from partitioned heritability | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#253](https://github.com/bulik/ldsc/issues/253) | ValueError: No objects to concatenate | Annotations/Partitioned | Provide a cts-annot walkthrough and validate that cts files align with ref LD and M counts. |
| [#252](https://github.com/bulik/ldsc/issues/252) | how to understand negative heritability or heritability values that are greater than 1 or even 2 | Regression Output/Interpretation | Explain negative estimates and when to treat h2/rg as unreliable; suggest checking sample size and QC. |
| [#251](https://github.com/bulik/ldsc/issues/251) | How to set N sample size on results from METAL while using munge_sumstats.py | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#249](https://github.com/bulik/ldsc/issues/249) | how to interpret genetic correction of two traits that are phenotypically correlated | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#245](https://github.com/bulik/ldsc/issues/245) | Issue with munge_sumstats.py : KeyError: u"None of [Int64Index([-1, -1, -1, -1, -1, -1, -1, -1, -1, -1,\n | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#244](https://github.com/bulik/ldsc/issues/244) | Warning message meaning : FutureWarning: `rcond` parameter will change to the default of machine precision times | Other/Uncategorized | Document the requested workflow or clarify constraints with a concrete example. |
| [#243](https://github.com/bulik/ldsc/issues/243) | Error for ldsc.py | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#242](https://github.com/bulik/ldsc/issues/242) | ValueError: Columns ['P'] are expected to be numeric | Input Format/Parsing | Handle extreme or non-numeric p-values (very small exponents, OR columns); provide guidance to clamp or convert. |
| [#241](https://github.com/bulik/ldsc/issues/241) | estimate_rg (FloatingPointError: invalid value encountered in sqrt) | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#240](https://github.com/bulik/ldsc/issues/240) | LDSC KeyError: "CHR" | Input Format/Parsing | Ensure required columns like `CHR` are present or allow column mapping; error should list required columns and example headers. |
| [#239](https://github.com/bulik/ldsc/issues/239) | Empty log files | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#237](https://github.com/bulik/ldsc/issues/237) | --overlap-annot does not work when reading in "thin" annot files? | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#236](https://github.com/bulik/ldsc/issues/236) | WARNING: median value of SIGNED_SUMSTATS is 0.5 (should be close to 0.0). This column may be mislabeled | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#234](https://github.com/bulik/ldsc/issues/234) | annotations with zero variance LD scores | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#233](https://github.com/bulik/ldsc/issues/233) | ValueError: LD Scores for concatenation must have identical SNP columns. | Input Format/Parsing | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#232](https://github.com/bulik/ldsc/issues/232) | Issue with new glob update | Other/Uncategorized | Document the requested workflow or clarify constraints with a concrete example. |
| [#231](https://github.com/bulik/ldsc/issues/231) | minor wget issue with the heritability example on the wiki | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#229](https://github.com/bulik/ldsc/issues/229) | heritability greater than 1 | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#228](https://github.com/bulik/ldsc/issues/228) | FloatingPointError: invalid value encountered in multiply | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#226](https://github.com/bulik/ldsc/issues/226) | What's Incompatible alleles mean? | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#224](https://github.com/bulik/ldsc/issues/224) | Manifest file for partitioned heritability | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#223](https://github.com/bulik/ldsc/issues/223) | Partitioned heritability: Error parsing reference panel LD Score. | Regression Output/Interpretation | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#222](https://github.com/bulik/ldsc/issues/222) | how to understand attenuation ratio? | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#220](https://github.com/bulik/ldsc/issues/220) | LD score regression trait correlation ERROR: FloatingPointError: invalid value encountered in sqrt | LD Score Computation/Performance | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#219](https://github.com/bulik/ldsc/issues/219) | Change of reference URL (Suggestion) | Reference Data/Downloads | Document stable reference file sources, versions, and expected contents; add preflight checks and mirror guidance for common downloads. |
| [#218](https://github.com/bulik/ldsc/issues/218) | Doubt about genetic correlation | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#214](https://github.com/bulik/ldsc/issues/214) | Error parsing refernce panel LD Score | Annotations/Partitioned | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#213](https://github.com/bulik/ldsc/issues/213) | Effect of --n-blocks on partitioning heritability | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#210](https://github.com/bulik/ldsc/issues/210) | doubt with sub-annotations - cell-type enrichment | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#208](https://github.com/bulik/ldsc/issues/208) | rg tutorial broken | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#207](https://github.com/bulik/ldsc/issues/207) | heritability obtain using --h2 and --rg are different. | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#206](https://github.com/bulik/ldsc/issues/206) | did not install it and have some errors below. | Install/Env | Provide zero-setup binaries/containers and precise dependency checks; surface clear diagnostics for missing system libs and avoid Python/conda/pandas issues. |
| [#205](https://github.com/bulik/ldsc/issues/205) | numpy.ufunc has the wrong size (MacOS 10.12.6) | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#204](https://github.com/bulik/ldsc/issues/204) | Continuous annotation for a subset of SNP: is it reasonable to set other SNP=0? | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#203](https://github.com/bulik/ldsc/issues/203) | Low ratio values - are the genetic correlation results reliable? | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#200](https://github.com/bulik/ldsc/issues/200) | Celltype/tissue specific analysis on ATAC seq data | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#195](https://github.com/bulik/ldsc/issues/195) | heritability estimation with summary data of mixed model | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#193](https://github.com/bulik/ldsc/issues/193) | Meaning of p-value in genetic correlation output (rg) | Regression Output/Interpretation | Handle extreme or non-numeric p-values (very small exponents, OR columns); provide guidance to clamp or convert. |
| [#192](https://github.com/bulik/ldsc/issues/192) | How to merge annotation files | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#191](https://github.com/bulik/ldsc/issues/191) | which N to use for large case/control imbalance data | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#190](https://github.com/bulik/ldsc/issues/190) | Make_annot.py issue and windows size to choose | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#189](https://github.com/bulik/ldsc/issues/189) | Genetic correlation significantly out of bounds [-1,1], with relatively small SE's | Regression Output/Interpretation | Explain why rg can exceed [-1,1] due to noise and provide interpretation guidance. |
| [#188](https://github.com/bulik/ldsc/issues/188) | R implementation | Other/Uncategorized | Document the requested workflow or clarify constraints with a concrete example. |
| [#187](https://github.com/bulik/ldsc/issues/187) | Munge error | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#183](https://github.com/bulik/ldsc/issues/183) | IOError could not open | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#181](https://github.com/bulik/ldsc/issues/181) | munge_sumstats.py never finished, then I had to use AWK | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#180](https://github.com/bulik/ldsc/issues/180) | numpy==1.16.0 !!! | Reference Data/Downloads | Document stable reference file sources, versions, and expected contents; add preflight checks and mirror guidance for common downloads. |
| [#179](https://github.com/bulik/ldsc/issues/179) | Tutorial data no longer available | Feature/UX | Add guidance or tooling for requested workflows (plots, tutorials, web resources) and clearly state what is out of scope. |
| [#178](https://github.com/bulik/ldsc/issues/178) | Anaconda3 does not work, really... | Install/Env | Emphasize the Rust binary removes Python/conda dependency issues; document minimal system requirements. |
| [#177](https://github.com/bulik/ldsc/issues/177) | Tag a release | Install/Env | Provide zero-setup binaries/containers and precise dependency checks; surface clear diagnostics for missing system libs and avoid Python/conda/pandas issues. |
| [#175](https://github.com/bulik/ldsc/issues/175) | KeyError: u"None of [Int64Index([-1, -2, -1, -2, -2, -1, -1, -1, -1, -1,\n | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#174](https://github.com/bulik/ldsc/issues/174) | FloatingPointError: invalid value encountered in less_equal | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#173](https://github.com/bulik/ldsc/issues/173) | pandas error | Install/Env | Emphasize the Rust binary removes Python/conda dependency issues; document minimal system requirements. |
| [#172](https://github.com/bulik/ldsc/issues/172) | X chromosome LD score regression | LD Score Computation/Performance | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#169](https://github.com/bulik/ldsc/issues/169) | local version vs LDhub | Install/Env | Provide zero-setup binaries/containers and precise dependency checks; surface clear diagnostics for missing system libs and avoid Python/conda/pandas issues. |
| [#167](https://github.com/bulik/ldsc/issues/167) | Estimate regions that contribute to enrichment | Annotations/Partitioned | Clarify annotation formats, baseline model expectations, and overlap rules; add validation that the annotation/LD score matrices align. |
| [#166](https://github.com/bulik/ldsc/issues/166) | LD-sc using chr:bp instead of rs numbers | Input Format/Parsing | Ensure required columns like `CHR` are present or allow column mapping; error should list required columns and example headers. |
| [#165](https://github.com/bulik/ldsc/issues/165) | Cell type specific analysis crashes with specific traits using a modified baseline model | Annotations/Partitioned | Explain baseline vs full baselineLD usage and provide a default recommended model. |
| [#161](https://github.com/bulik/ldsc/issues/161) | error in running munge_sumstats.py | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#160](https://github.com/bulik/ldsc/issues/160) | Error in munge_sumstats.py if multiple columns have a "default name" | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#159](https://github.com/bulik/ldsc/issues/159) | munge_sumstats.py error | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#158](https://github.com/bulik/ldsc/issues/158) | LinAlgError: Singular matrix | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#157](https://github.com/bulik/ldsc/issues/157) | KeyError in munge_sumstats step | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#156](https://github.com/bulik/ldsc/issues/156) | Prop._h2 outside [0, 1] | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#155](https://github.com/bulik/ldsc/issues/155) | negative h2 and Ratio > 1 | Regression Output/Interpretation | Explain negative estimates and when to treat h2/rg as unreliable; suggest checking sample size and QC. |
| [#154](https://github.com/bulik/ldsc/issues/154) | minimum sequence or number of SNPs needed for estimating partitioned heritability in my own annotations? | Regression Output/Interpretation | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#153](https://github.com/bulik/ldsc/issues/153) | Issue related to method derivation | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#152](https://github.com/bulik/ldsc/issues/152) | -interecept-gencov | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#151](https://github.com/bulik/ldsc/issues/151) | Heritability become less after imputation | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#150](https://github.com/bulik/ldsc/issues/150) | munge_sumstats.py throwing error on simulated dataset | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#149](https://github.com/bulik/ldsc/issues/149) | Estimate heritability | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#148](https://github.com/bulik/ldsc/issues/148) | Tutorial Cell type specific analyses fails | Feature/UX | Add guidance or tooling for requested workflows (plots, tutorials, web resources) and clearly state what is out of scope. |
| [#147](https://github.com/bulik/ldsc/issues/147) | Cell type analysis: IOError: Could not open Cahoy.1.l2.ldscore[./gz/bz2] | LD Score Computation/Performance | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#145](https://github.com/bulik/ldsc/issues/145) | FIXED - Fail to converst summary statistic in .sumstats format: munge_sumstats is taking hours | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#144](https://github.com/bulik/ldsc/issues/144) | Error converting summary statistics UKBiobank p-values | Input Format/Parsing | Handle extreme or non-numeric p-values (very small exponents, OR columns); provide guidance to clamp or convert. |
| [#143](https://github.com/bulik/ldsc/issues/143) | Can i use ldsc to calculate quantitative traits. | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#141](https://github.com/bulik/ldsc/issues/141) | Questions about genesets' sample size when doing the regression | Other/Uncategorized | Document the requested workflow or clarify constraints with a concrete example. |
| [#140](https://github.com/bulik/ldsc/issues/140) | Heritability have error | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#139](https://github.com/bulik/ldsc/issues/139) | Running Error | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#138](https://github.com/bulik/ldsc/issues/138) | Prop._h2 always is 1.0 use my own ldscores based on my own annot | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#136](https://github.com/bulik/ldsc/issues/136) | [Cell-type specific analysis] How are the 'control' gene sets determined? | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#135](https://github.com/bulik/ldsc/issues/135) | Include 'P-value' | Input Format/Parsing | Handle extreme or non-numeric p-values (very small exponents, OR columns); provide guidance to clamp or convert. |
| [#132](https://github.com/bulik/ldsc/issues/132) | Error running munge_sumstats.py | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#131](https://github.com/bulik/ldsc/issues/131) | Error converting summary statistics | Other/Uncategorized | Document the requested workflow or clarify constraints with a concrete example. |
| [#125](https://github.com/bulik/ldsc/issues/125) | Error converting summary statistic | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#123](https://github.com/bulik/ldsc/issues/123) | SNP-heritability calculation requires consecutive chromosome number? | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#122](https://github.com/bulik/ldsc/issues/122) | Threads used when estimating LD scores | LD Score Computation/Performance | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#121](https://github.com/bulik/ldsc/issues/121) | ERROR converting summary statistics | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#117](https://github.com/bulik/ldsc/issues/117) | LD score regression trait correlation ERROR:invalid value encountered in sqrt | LD Score Computation/Performance | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#116](https://github.com/bulik/ldsc/issues/116) | Incomplete data - A1/A2 column is not available | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#114](https://github.com/bulik/ldsc/issues/114) | Incorrect SNP filtering and MAF when using --keep in LD score calculations | LD Score Computation/Performance | Explain expected MAF/LD correlation ranges and note that small negatives can occur due to noise. |
| [#112](https://github.com/bulik/ldsc/issues/112) | ./munge_sumstats.py is running for a long time | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#111](https://github.com/bulik/ldsc/issues/111) | To analyze a new annotation, you can add it to the full baseline model. | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#110](https://github.com/bulik/ldsc/issues/110) | partitioning heritability | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#109](https://github.com/bulik/ldsc/issues/109) | how to understand the results of partitioning heritability | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#107](https://github.com/bulik/ldsc/issues/107) | LD scores for indels from 1000 genome | LD Score Computation/Performance | Explain LD window choices, performance tradeoffs, and add warnings when the window spans a full chromosome. |
| [#105](https://github.com/bulik/ldsc/issues/105) | Mixed delimiters in header of baselineLD_v1.1 annotation | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#104](https://github.com/bulik/ldsc/issues/104) | Error converting summary statistics | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#103](https://github.com/bulik/ldsc/issues/103) | Run Partitioned heritability have errors | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#102](https://github.com/bulik/ldsc/issues/102) | Partitioned heritability by functional category | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#99](https://github.com/bulik/ldsc/issues/99) | Discrepancies in output (results table) for Liability Scale Heritability | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#97](https://github.com/bulik/ldsc/issues/97) | Partitioned heritability inflated SE and FloatingPointError | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#96](https://github.com/bulik/ldsc/issues/96) | Baseline Model for Partitioning Analysis | Annotations/Partitioned | Explain baseline vs full baselineLD usage and provide a default recommended model. |
| [#94](https://github.com/bulik/ldsc/issues/94) | Genetic Correlation: nan (nan) (h2 out of bounds) | Regression Output/Interpretation | Explain why rg can exceed [-1,1] due to noise and provide interpretation guidance. |
| [#93](https://github.com/bulik/ldsc/issues/93) | Errors running | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#92](https://github.com/bulik/ldsc/issues/92) | Z-scores New Categories | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#91](https://github.com/bulik/ldsc/issues/91) | Partitioned heritability | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#90](https://github.com/bulik/ldsc/issues/90) | error launching munge_sumstats.py | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#89](https://github.com/bulik/ldsc/issues/89) | why are the genetic correlation out of the bound of [-1, 1] and have very large se? | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#88](https://github.com/bulik/ldsc/issues/88) | why the estimated heritablity is negative? | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#87](https://github.com/bulik/ldsc/issues/87) | The value after --intercept-gencov flag is the correct intercept term | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#85](https://github.com/bulik/ldsc/issues/85) | Unalignable boolean Series provided as indexer | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#84](https://github.com/bulik/ldsc/issues/84) | Question about partition heritability stratified by allele freqency | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#83](https://github.com/bulik/ldsc/issues/83) | Issues with ldscores from 2 LDSC_Seg annotations | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#82](https://github.com/bulik/ldsc/issues/82) | about the p value of genetic correlation for GC correction GWAS | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#79](https://github.com/bulik/ldsc/issues/79) | Feature Request | Feature/UX | Add guidance or tooling for requested workflows (plots, tutorials, web resources) and clearly state what is out of scope. |
| [#77](https://github.com/bulik/ldsc/issues/77) | Error while partitioning heritability with the baseline model | Regression Output/Interpretation | Explain baseline vs full baselineLD usage and provide a default recommended model. |
| [#68](https://github.com/bulik/ldsc/issues/68) | How can I get the partitioned heritability by chromosome 1-22 other than functional category? | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#66](https://github.com/bulik/ldsc/issues/66) | Error converting summary statistics | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#64](https://github.com/bulik/ldsc/issues/64) | Errors in tests with nosetests | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#58](https://github.com/bulik/ldsc/issues/58) | Validate expected dtypes in munge_sumstats | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#43](https://github.com/bulik/ldsc/issues/43) | fatal error while running munge_sumstats.py | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#34](https://github.com/bulik/ldsc/issues/34) | error for generate ldscore file when no annotation in that chromosome | Annotations/Partitioned | Validate annotation dimensions vs .bim and document the exact .annot/.annot.gz formats. |
| [#32](https://github.com/bulik/ldsc/issues/32) | infer N from beta [SE or P] and reference panel MAF | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |
| [#25](https://github.com/bulik/ldsc/issues/25) | better error messages for --rg | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#24](https://github.com/bulik/ldsc/issues/24) | deal with mismatched alleles in sumstats files not generated with --merge-alleles | Input Format/Parsing | Document `--merge-alleles` format (SNP, A1, A2) and show how allele mismatches are handled. |
| [#20](https://github.com/bulik/ldsc/issues/20) | warn if rg driven by single locus | Regression Output/Interpretation | Explain common result pathologies (NA/negative/rg>1) and add diagnostics so users know when estimates are unreliable. |
| [#12](https://github.com/bulik/ldsc/issues/12) | document .frq file format | Input Format/Parsing | Validate inputs early with explicit error messages (missing columns, bad gzip, allele mismatches); offer column mapping examples and strict/lenient modes. |