DeepMapper¶

No-filtering, attribute-back analysis of high-dimensional omics.

DeepMapper keeps every feature (no highly-variable-gene selection, no dimension reduction), reads the full matrix with a configurable backbone over several feature arrangements, and attributes each result back to named features. That surfaces distributed gene chords, sets of genes that separate a cell state only together, which standard pipelines discard. An optional de-novo step recovers transcripts absent from the reference annotation.

Install¶

pip install pydeepmapper            # core engine
pip install "pydeepmapper[all]"     # + io, backbones, de-novo

Quickstart¶

from pydeepmapper.config import DeepMapperConfig, BackboneSpec
from pydeepmapper.runner import run

cfg = DeepMapperConfig(n_passes=3, backbone=BackboneSpec(kind="cnn_small"))
findings = run(X, y, cfg, feature_names=genes)   # X unfiltered, y integer labels
for name, freq, importance in findings.ranking(20):
    print(name, round(freq, 3), round(importance, 4))

Where to go next¶

User manual is the full guide: configuration, data loading, evaluation, attribution, early stopping, de-novo recovery, package layout.
API reference is generated from the docstrings.
Reproducing the analyses maps every figure to its script.
Using DeepMapper with Claude drives an analysis through the bundled Claude Code skill.

Citation¶

Ersavas T., Smith M.A., Mattick J.S. (2024). Novel applications of Convolutional Neural Networks in the age of Transformers. Scientific Reports 14. https://doi.org/10.1038/s41598-024-60709-z

To cite this software release, use the Zenodo archive: https://doi.org/10.5281/zenodo.20967454