bio-expression-matrix-counts-ingest

Version Compatibility

Reference examples tested with: pandas 2.2+

Before using code patterns, verify installed versions match. If versions differ:

• Python: pip show <package> then help(module.function) to check signatures
• R: packageVersion('<pkg>') then ?function_name to verify parameters

If code throws ImportError, AttributeError, or TypeError, introspect the installed package and adapt the example to match the actual API rather than retrying.

Count Matrix Ingestion

Basic CSV/TSV Loading

Goal: Load a gene expression count matrix from a delimited text file into a pandas DataFrame.

Approach: Read CSV/TSV with gene IDs as the row index, handling comment lines if present.

"Load my count matrix" → Read a delimited file into a DataFrame with genes as rows and samples as columns.

import pandas as pd

# TSV with gene IDs as first column
counts = pd.read_csv('counts.tsv', sep='\t', index_col=0)

# CSV with header
counts = pd.read_csv('counts.csv', index_col=0)

# Skip comment lines
counts = pd.read_csv('counts.txt', sep='\t', index_col=0, comment='#')

featureCounts Output

Goal: Parse featureCounts output into a clean count matrix by stripping metadata columns.

Approach: Skip the 6 annotation columns (Chr, Start, End, Strand, Length) and clean BAM path suffixes from column names.

import pandas as pd

# featureCounts format has 6 metadata columns before counts
fc = pd.read_csv('featurecounts.txt', sep='\t', comment='#')
counts = fc.set_index('Geneid').iloc[:, 5:]  # Skip Chr, Start, End, Strand, Length
counts.columns = [c.replace('.bam', '').split('/')[-1] for c in counts.columns]

Salmon Quant Files

Goal: Combine per-sample Salmon quantification files into a single count or TPM matrix.

Approach: Iterate over quant directories, extract the desired column from each quant.sf, and merge into a DataFrame.

import pandas as pd
from pathlib import Path

def load_salmon_quants(quant_dirs, column='NumReads'):
    '''Load multiple Salmon quant.sf files into a count matrix.'''
    dfs = {}
    for qdir in quant_dirs:
        sample = Path(qdir).name
        sf = pd.read_csv(f'{qdir}/quant.sf', sep='\t', index_col=0)
        dfs[sample] = sf[column]
    return pd.DataFrame(dfs)

# Usage
quant_dirs = ['salmon_out/sample1', 'salmon_out/sample2', 'salmon_out/sample3']
counts = load_salmon_quants(quant_dirs, column='NumReads')
tpm = load_salmon_quants(quant_dirs, column='TPM')