Data Science & Machine Learning for Biology – Tiny Lab BioML

From Fingerprints to Message Passing: Benchmarking Three Approaches to Molecular Toxicity Prediction on Tox21

Python

Deep Learning

Graph Neural Network

PyTorch

Cheminformatics

Machine Learning

In this post, I compare three approaches to predicting chemical toxicity directly from molecular structure, using the Tox21 dataset (~7,800 compounds tested across 12…

Querying ChEMBL for Target-Specific Bioactivity and Compound Data in Python

Python

ChEMBL

Cheminformatics

Small, reusable Python utilities for pulling target-specific bioactivity and compound data from ChEMBL given a UniProt ID, and evaluating those compounds against Lipinski’s…

Diagnosing Overfitting in a PPI Transformer: Hyperparameter Search and What Actually Helped

Python

Deep Learning

Transformer

PyTorch

ESM2

Protein Language Models

A hyperparameter search was conducted to diagnose and mitigate overfitting in a transformer-based model for predicting Protein-Protein Interactions (PPIs). The search…

Predicting Protein-Protein Interactions with ESM2 Embeddings and Transformer-based Interaction Model

Python

Deep Learning

Transformer

PyTorch

ESM2

Protein Language Models

Data Leakage

Using the PyTorch framework, I built a transformer-based model to predict Protein-Protein Interactions (PPIs) trained from the HuRI dataset. The workflow takes a pair of…

Evaluating Data Leakage in Protein Binding Affinity Prediction

Python

Deep Learning

Convolutional Neural Networks

Self-Attention

Keras

ESM2

Protein Language Models

Data Leakage

Data leakage is a critical issue in machine learning, especially in the context of protein-protein interaction (PPI) prediction. Data leakage occurs when information from…

Peptide-MHCII Binding Affinity Prediction with Convolutional Neural Networks and Protein Large Language Model

Python

Deep Learning

Convolutional Neural Networks

Keras

ESM2

Protein Language Models

To accurately predict peptide-MHCII binding affinity, I developed a 1D Convolutional Neural Network (CNN) combined with Multi-Layer Perceptron (MLP) model using protein…

Predicting Peptide-MHC Class II Binding with ESM2 and Neural Networks

Python

Deep Learning

Keras

ESM2

Protein Language Models

In this post, I will build a neural network to predict the binding affinity between peptides and MHC Class II alleles, utilizing embeddings extracted from the Evolutionary…

Machine learning for drug sensitivity prediction (Part 3): Training an Elastic Net gene expression model to predict Erlotinib sensitivity

R

DepMap

Machine Learning

Elastic Net

Caret

In my previous posts (part 1 and part 2), I explored whether using autoencoders (AE) to compress high-dimensional multi-omics data could improve the performance of drug…

Training a Random Forest model with Scikit-Learn on DepMap data

Python

Machine Learning

DepMap

Random Forest

Scikit-Learn

Pandas

In this specific blog post, we will explore the Cancer Dependency Map (DepMap) dataset to understand the relationship between genetic dependencies (CRISPR knockout effects)…

Machine learning for drug sensitivity prediction (Part 2): will deep features improve accuracy?

R

DepMap

Autoencoder

Multi-omics

Machine Learning

Random Forest

Elastic Net

Caret

In Part 1, I explored using autoencoders (AE) to compress high-dimensional DepMap omics data into deep features. I visualized these features using UMAP and observed that…

Machine learning for drug sensitivity prediction (Part 1): using autoencoders to extract deep features from DepMap OMICS

R

DepMap

Autoencoder

Multi-omics

Machine Learning

During my time in the industry, I have analyzed quite a few cell panel screens aiming to identify biomarkers of drug sensitivity. These screens are often carried out by…