biors 0.28.0

Command-line tools for bio-rs biological AI model input workflows.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131

use crate::errors::CliError;
use crate::output::print_success;
use biors_core::{ModelInputRecord, SequenceWorkflowOutput, TokenizedProtein, ValidatedSequence};
use serde::Serialize;
use std::collections::BTreeMap;

pub(crate) fn write_debug(output: SequenceWorkflowOutput) -> Result<(), CliError> {
    let debug = SequenceDebugOutput::from_workflow(&output);
    print_success(Some(output.provenance.input_hash), debug)
}

#[derive(Debug, Serialize)]
struct SequenceDebugOutput {
    view: &'static str,
    records: Vec<SequenceDebugRecord>,
}

#[derive(Debug, Serialize)]
struct SequenceDebugRecord {
    id: String,
    normalized_sequence: String,
    token_map: Vec<TokenDebugStep>,
    model_input: Option<ModelInputRecord>,
    error_visualization: ErrorVisualization,
}

#[derive(Debug, Serialize)]
struct TokenDebugStep {
    position: usize,
    residue: char,
    token_id: u8,
    status: &'static str,
}

#[derive(Debug, Serialize)]
struct ErrorVisualization {
    sequence: String,
    markers: String,
    legend: &'static str,
}

impl SequenceDebugOutput {
    fn from_workflow(workflow: &SequenceWorkflowOutput) -> Self {
        let model_records: BTreeMap<_, _> = workflow
            .model_input
            .as_ref()
            .map(|input| {
                input
                    .records
                    .iter()
                    .map(|record| (record.id.as_str(), record.clone()))
                    .collect()
            })
            .unwrap_or_default();

        let records = workflow
            .validation
            .sequences
            .iter()
            .zip(workflow.tokenization.records.iter())
            .map(|(validated, tokenized)| SequenceDebugRecord {
                id: validated.id.clone(),
                normalized_sequence: validated.sequence.clone(),
                token_map: token_debug_steps(validated, tokenized),
                model_input: model_records.get(validated.id.as_str()).cloned(),
                error_visualization: error_visualization(validated),
            })
            .collect();

        Self {
            view: "sequence_debug.v0",
            records,
        }
    }
}

fn token_debug_steps(
    validated: &ValidatedSequence,
    tokenized: &TokenizedProtein,
) -> Vec<TokenDebugStep> {
    validated
        .sequence
        .chars()
        .enumerate()
        .map(|(index, residue)| {
            let position = index + 1;
            TokenDebugStep {
                position,
                residue,
                token_id: tokenized.tokens.get(index).copied().unwrap_or_default(),
                status: token_status(position, validated),
            }
        })
        .collect()
}

fn token_status(position: usize, validated: &ValidatedSequence) -> &'static str {
    if validated
        .errors
        .iter()
        .any(|issue| issue.position == position)
    {
        "error"
    } else if validated
        .warnings
        .iter()
        .any(|issue| issue.position == position)
    {
        "warning"
    } else {
        "standard"
    }
}

fn error_visualization(validated: &ValidatedSequence) -> ErrorVisualization {
    let markers: String = validated
        .sequence
        .chars()
        .enumerate()
        .map(|(index, _)| match token_status(index + 1, validated) {
            "error" => 'E',
            "warning" => 'W',
            _ => '.',
        })
        .collect();
    ErrorVisualization {
        sequence: validated.sequence.clone(),
        markers,
        legend: ". standard, W warning, E error",
    }
}