ProteomeLM-L-leprae essentiality predictions

HANSEN Essentiality Prediction

Proteome-wide predictions of gene essentiality in Mycobacterium leprae, identifying the proteins the bacterium is most likely to require for survival. Predictions were generated by ProteomeLM-L-leprae, a classifier trained on transposon mutagenesis measurements of essentiality in M. tuberculosis and transferred to M. leprae. Each protein carries a calibrated probability of essentiality together with one of four classes, so that proteins can be ranked consistently across the proteome. Where a one-to-one M. tuberculosis orthologue exists, its experimentally determined essentiality is reported alongside the prediction.

Conservative interpretation

Proteins are assigned to classes with deliberately conservative boundaries, so that weakly supported predictions are not over-interpreted.

Proteins

—

Essential

—

Likely Essential

—

Uncertain

—

Non‑essential

—

Per-protein ProteomeLM-L-leprae Essentiality

Proteins are assigned to one of four essentiality classes according to their calibrated probability, with class boundaries at 0.315, 0.222 and 0.165. The classifier was trained on transposon mutagenesis measurements of essentiality in M. tuberculosis and transferred to M. leprae. Where a one-to-one M. tuberculosis orthologue exists, its experimentally determined essentiality is reported alongside the prediction.

Download TSV
Rank ? ML ID ? Gene Protein Essentiality ? ProteomeLM‑L‑leprae ? Mtb Anchor ?
Loading…

How the predictions were generated

Essentiality was predicted with a classifier trained on laboratory measurements of gene essentiality in M. tuberculosis (PMID: 28096490) and applied to M. leprae, where it ranked essential genes above non-essential ones in 84% of pairs. The calibrated probabilities are best used to rank and prioritise proteins rather than as exact estimates, and where a one-to-one M. tuberculosis orthologue exists its experimentally determined essentiality is shown for comparison.

Read methods in Help