Dan Ofer

DNT: Diploid Genomic Foundation Model

The paper introduces a diploid genomic foundation model to represent diploid genotypes in genomic language models, improving variant discrimination in compound-heterozygous cases.

OTRec: Deep learning recommender for prospective druggable disease-target associations

OTRec is a deep learning model that recommends druggable disease-target associations, outperforming existing methods in both retrospective and cross-validation settings.

GFMBench-API: A Standardized Interface for Benchmarking Genomic Foundation Models

Wrap a genomic foundation model's inference methods once to score it on supervised and zero-shot GUE, VariantBenchmarks, BEND, TraitGym and ClinVar tasks; code is promised upon acceptance.

On the state of protein function prediction: a report on the fourth CAFA challenge

This paper reports on the fourth CAFA challenge, showing improvements in Gene Ontology term prediction but highlighting challenges in new evaluation settings.

MAGI: Mechanistic Consequences of Genetic Variants via Genomic Foundation Models

MAGI provides mechanistic insights into genetic variants by integrating genomic data and clinical annotations to improve variant interpretation and generate testable hypotheses.

Improving Causal Gene Identification Using Large Language Models

This study evaluates using Large Language Models for causal gene identification, showing improvements with Retrieval-Augmented Generation and genomic distance information, but with diminishing returns when combined.

Protein Language Models Expose Viral Mimicryand Immune Escape

The paper uses protein language models to distinguish human and viral proteins, identifying immune escape mechanisms through error analysis.

Detecting Anomalous Proteins Using Deep Representations

This paper uses deep learning to detect anomalous proteins by identifying unusual properties in function, phylogenetic families, and sequence segmentation.

Inferring microRNA regulation: A proteome perspective

The paper predicts miRNA-regulated genes using a protein-based machine learning framework, showing better performance than genomic approaches and identifying exceptions like GPCRs.

ProteinBERT: A universal deep-learning model of protein sequence and function

A ~16M-parameter protein language model, pretrained on ~106M UniRef90 proteins by denoising sequences and Gene Ontology annotations, rivals larger TAPE models using linear-cost global attention.

ASAP: A Machine-Learning Framework for Local Protein Properties

ASAP is a machine-learning framework for predicting residue-level protein properties, demonstrating high performance in predicting protein cleavage sites.