Struct SeqStats 

pub struct SeqStats {
    pub count: u64,
    pub total_length: u64,
    pub min_length: Option<u64>,
    pub max_length: Option<u64>,
    pub bases: BaseCounts,
    /* private fields */
}

Accumulates length, composition and quality statistics.

use fastx::{FastxReader, SeqStats};

let data = b">a\nACGTACGTAC\n>b\nGGCC\n";
let mut stats = SeqStats::default();
FastxReader::new(&data[..]).for_each_record(|r| { stats.push(r); Ok(()) })?;

assert_eq!(stats.count, 2);
assert_eq!(stats.total_length, 14);
assert_eq!(stats.min_length, Some(4));
assert_eq!(stats.max_length, Some(10));
assert_eq!(stats.n50(), Some(10));

Fields

count: u64

Number of records seen.

total_length: u64

Sum of all sequence lengths.

min_length: Option<u64>

Shortest record, if any.

max_length: Option<u64>

Longest record, if any.

bases: BaseCounts

Base composition across all records.

Implementations

impl SeqStats

pub fn new() -> SeqStats

An empty accumulator.

pub fn push(&mut self, record: &Sequence)

Fold one record in.

pub fn merge(&mut self, other: &SeqStats)

Merge another accumulator, for parallel or per-file aggregation.

pub fn is_empty(&self) -> bool

True when no records have been seen.

pub fn mean_length(&self) -> Option<f64>

Mean sequence length.

pub fn median_length(&self) -> Option<u64>

Median sequence length (lower median for an even count).

pub fn n50(&self) -> Option<u64>

N50: contigs of at least this length cover half the total.

pub fn n90(&self) -> Option<u64>

N90.

pub fn l50(&self) -> Option<u64>

L50: the number of contigs that make up the N50.

pub fn gc_content(&self) -> Option<f64>

GC fraction over unambiguous bases.

pub fn quality_bases(&self) -> u64

Number of quality characters seen across all records.

pub fn quality_histogram(&self) -> &[u64; 94]

Counts per Phred score, indexed by score — the same data FastQC plots.

let mut stats = SeqStats::new();
stats.push(&Sequence::fastq("r", b"ACGT", b"IIII")?);
assert_eq!(stats.quality_histogram()[40], 4); // 'I' is Q40

pub fn expected_errors(&self) -> f64

Expected number of wrong bases across all records.

pub fn mean_quality(&self) -> Option<f64>

Mean quality as a Phred score, derived from the mean error rate.

pub fn q20_fraction(&self) -> Option<f64>

Fraction of bases at Q20 or better.

pub fn q30_fraction(&self) -> Option<f64>

Fraction of bases at Q30 or better.

pub fn fraction_at_least(&self, score: u8) -> Option<f64>

Fraction of bases whose Phred score is at least score.

pub fn lengths(&self) -> &[u64]

All observed lengths, in the order the records were seen.

pub fn to_json(&self) -> String

Render the statistics as a JSON object, for pipelines that parse output.

The object is on one line, so a run over several files is valid line-delimited JSON and streams straight into jq. Pipe through jq . if you want it laid out.

Absent figures — quality for FASTA, N50 after SeqStats::forget_lengths — come out as null rather than being omitted, so the shape of the object never changes.

let mut stats = SeqStats::new();
stats.push(&Sequence::fasta("a", b"ACGT"));
let json = stats.to_json();
assert!(json.starts_with('{') && json.ends_with('}'));
assert!(!json.contains('\n'), "must stay on one line");
assert!(json.contains("\"records\":1"));
assert!(json.contains("\"mean_quality\":null"));

pub fn forget_lengths(&mut self)

Drop the retained per-record lengths to cap memory on huge inputs.

After this call SeqStats::n50, SeqStats::median_length and SeqStats::l50 return None, but counts and means stay correct.

Trait Implementations

impl Clone for SeqStats

fn clone(&self) -> SeqStats

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for SeqStats

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for SeqStats

fn default() -> SeqStats

Returns the “default value” for a type.

impl Display for SeqStats

fn fmt(&self, f: &mut Formatter<'_>) -> Result

A block of aligned label value lines, without a trailing newline.