Genomic Pre-trained Network

Tissue-specific variant effect predictions on splicing

Fully convolutional deep learning variant effect predictor architecture

Predicting the effect of mutations on protein stability and protein binding affinity using pretrained neural networks and a ranking objective function.

Clinical Whole Genome and Exome Sequencing Pipeline

We present Envision, an accurate predictor of protein variant molecular effect, trained using large-scale experimental mutagenesis data. All data and software in this study are freely available. The training data set and all code used to train the models and generate the figures presented in this manuscript are available here. Envision predictio…

CADD-SV – a framework to score the effect of structural variants

Implementation of SpliceAI, Illumina's deep neural network to predict variant effects on splicing, in PyTorch.

Pipeline for variant annotation using Variant Effect Predictor (VEP)

Implementation of evolutionary model of variant effect (EVE), a deep generative model of evolutionary data, in PyTorch.

A Haskell script that performs basic parsing on the default output of ensembl-vep and variant-calling format (vcf) files.

Preliminary analysis of a combined DMS dataset, including clustering and VEP benchmark

Deep learning model for non-coding regulatory variants

Deep learning framework to predict functional effects of missense variants in human

Master's Thesis Project

interactive variant tables for easy filtering

An explorative tool to identify functional hotspots

Transcriptional variant verification to validate predicted variants from genomic data in expressed transcripts

Experimenting with protein language model predictions