Alexander Kitaygorodsky

AtomWeaver: Multi-Component Flow Matching with a Structured Geometric Prior Facilitates Non-Canonical Peptide Design

AtomWeaver enables non-canonical peptide design by using a geometric prior to generate side-chain atoms without restricting residue identity, achieving high agreement with experimental results.

Improved De Novo Peptide Binder Design with Target-Conditioned Inverse Folding

This paper improves de novo peptide binder design by combining peptide-specific finetuning with a novel decoding order strategy, achieving a 16-30% increase in in silico success rates and doubling top-ranking agonists for OPRM1.

Predicting localized affinity of RNA binding proteins to transcripts with convolutional neural networks

The paper introduces POLARIS, a deep-learning method for predicting localized RBP binding affinity to transcripts, achieving high performance across cell lines and enabling interpretation of noncoding variants.

Whole Genome De Novo Variant Identification with FreeBayes and Neural Network Approaches

The paper develops a framework for de novo variant identification using FreeBayes and a neural network, achieving high accuracy in SNVs and indels through validation with Sanger sequencing.

Genetic analysis of de novo variants reveals sex differences in complex and isolated congenital diaphragmatic hernia and indicates MYRF as a candidate gene

The paper analyzes de novo variants in congenital diaphragmatic hernia cases to identify sex differences and a new candidate gene MYRF.