Skip to content

Home

Home

← Back to collection

Share this slide deck

Email LinkedIn WhatsApp Bluesky X Facebook

← Prev 17/87 Next →

Student conference

Exploring the Application of Protein Language Models in Lung Adenocarcinoma Survival Analysis

By Matthew Santos

EMS300 Student Conference 2026

Slide 1: Two patients. Same cancer.. A full description follows.

1 of 11

Summary

The application of Protein Language Models (PLMs) in predicting survival outcomes for lung adenocarcinoma, a leading cause of cancer mortality, is explored. Current survival predictions rely heavily on TNM staging and binary mutation classifications, which overlook critical biological variations. It is hypothesised that attention-weighted ESM2 embeddings can capture meaningful differences in missense mutations, potentially enhancing survival predictions. The study utilises data from 448 lung cancer patients, comparing various survival models to assess the impact of incorporating PLM-derived data.

PeerMosaic

Academic work, brought together.