LLMs-Universal-Life-Science-and-Clinical-Skills- mixcr-analysis

<!--

install

source · Clone the upstream repo

git clone https://github.com/mdbabumiamssm/LLMs-Universal-Life-Science-and-Clinical-Skills-

Claude Code · Install into ~/.claude/skills/

T=$(mktemp -d) && git clone --depth=1 https://github.com/mdbabumiamssm/LLMs-Universal-Life-Science-and-Clinical-Skills- "$T" && mkdir -p ~/.claude/skills && cp -r "$T/Skills/Immunology_Vaccines/bioSkills/mixcr-analysis" ~/.claude/skills/mdbabumiamssm-llms-universal-life-science-and-clinical-skills-mixcr-analysis && rm -rf "$T"

manifest: Skills/Immunology_Vaccines/bioSkills/mixcr-analysis/SKILL.md

source content

name: bio-tcr-bcr-analysis-mixcr-analysis description: Perform V(D)J alignment and clonotype assembly from TCR-seq or BCR-seq data using MiXCR. Use when processing raw immune repertoire sequencing data to identify clonotypes and their frequencies. tool_type: cli primary_tool: MiXCR measurable_outcome: Execute skill workflow successfully with valid output within 15 minutes. allowed-tools:

read_file
run_shell_command

MiXCR Analysis

Complete Workflow (Recommended)

mixcr analyze generic-tcr-amplicon \
    --species human \
    --rna \
    --rigid-left-alignment-boundary \
    --floating-right-alignment-boundary C \
    input_R1.fastq.gz input_R2.fastq.gz \
    output_prefix

mixcr analyze 10x-vdj-tcr \
    input_R1.fastq.gz input_R2.fastq.gz \
    output_prefix

Step-by-Step Workflow

Step 1: Align Reads

mixcr align \
    --species human \
    --preset generic-tcr-amplicon-umi \
    input_R1.fastq.gz input_R2.fastq.gz \
    alignments.vdjca

mixcr align \
    --species human \
    --rna \
    -OallowPartialAlignments=true \
    input_R1.fastq.gz input_R2.fastq.gz \
    alignments.vdjca

Step 2: Refine and Assemble

mixcr refineTagsAndSort alignments.vdjca alignments_refined.vdjca

mixcr assemble alignments_refined.vdjca clones.clns

Step 3: Export Results

mixcr exportClones \
    --chains TRB \
    --preset full \
    clones.clns \
    clones.tsv

mixcr exportClones \
    --chains TRB \
    -cloneId -readCount -readFraction \
    -nFeature CDR3 -aaFeature CDR3 \
    -vGene -dGene -jGene \
    clones.clns \
    clones_custom.tsv

Preset Protocols

Protocol	Use Case
`generic-tcr-amplicon`	TCR amplicon sequencing
`generic-bcr-amplicon`	BCR amplicon sequencing
`generic-tcr-amplicon-umi`	TCR amplicon with UMIs
`rnaseq-tcr`	TCR extraction from bulk RNA-seq
`rnaseq-bcr`	BCR extraction from bulk RNA-seq
`10x-vdj-tcr`	10x Genomics TCR enrichment
`10x-vdj-bcr`	10x Genomics BCR enrichment
`takara-human-tcr-v2`	Takara SMARTer kit

Species Support

mixcr align --species human ...
mixcr align --species mmu ...

# Available: human, mmu, rat, rhesus, dog, pig, rabbit, chicken

Output Format

Column	Description
cloneId	Unique clone identifier
readCount	Number of reads
cloneFraction	Proportion of repertoire
nSeqCDR3	Nucleotide CDR3 sequence
aaSeqCDR3	Amino acid CDR3 sequence
allVHitsWithScore	V gene assignments
allDHitsWithScore	D gene assignments
allJHitsWithScore	J gene assignments

Quality Metrics

mixcr exportReports alignments.vdjca

# Key metrics:
# - Successfully aligned reads (>80% is good)
# - CDR3 found (>70% of aligned)
# - Clonotype count (varies by sample type)

Parse MiXCR Output in Python

import pandas as pd

def load_mixcr_clones(filepath):
    df = pd.read_csv(filepath, sep='\t')
    df = df.rename(columns={
        'readCount': 'count',
        'cloneFraction': 'frequency',
        'aaSeqCDR3': 'cdr3_aa',
        'nSeqCDR3': 'cdr3_nt'
    })
    return df

Related Skills

vdjtools-analysis - Downstream diversity analysis
scirpy-analysis - Single-cell VDJ integration
repertoire-visualization - Visualize MiXCR output