Sam Boshar

Sam Boshar builds genomic foundation models at InstaDeep, working on sequence modeling, protein-task applications, controllable generation, and long-context genomic prediction. He previously co-authored research evaluating genomic language models on protein tasks and completed an MIT master’s thesis on genomic language models for protein function and property prediction.

Career

Cosigned byLoading...

CosignedSign in to view

Based in: New York City

InstaDeep

Machine Learning Engineer
Jul 2024 — Present
Cambridge, Massachusetts, United States
Research on genomic language models, protein downstream tasks, controllable sequence generation and long-context genomic foundation models.
Machine Learning Research Co-op @ Instadeep
Jun 2023 — May 2024
Boston, Massachusetts, United States

Massachusetts Institute of Technology

MEng

Massachusetts Institute of Technology

BS, Computer Science

Phillips Academy

Products & media

instadeep.com

Nucleotide Transformer v3

Writing

pmc.ncbi.nlm.nih.gov

Are genomic language models all you need?

Writing

dspace.mit.edu

Genomic Language Models for Protein Function and Property Prediction

His MIT thesis investigates genomic language models for protein prediction, including how genomic and protein representations can complement one another.

Writing

Profile name

Company name

Company name

Profile name

Company name

Company name

Profile name

Company name

Company name

Profile name

Company name

Company name

Profile name

Company name

Company name

Sam Boshar

InstaDeep

Machine Learning Engineer

Machine Learning Research Co-op @ Instadeep

Massachusetts Institute of Technology

Massachusetts Institute of Technology

Phillips Academy

Products & media

Nucleotide Transformer v3

Are genomic language models all you need?

Genomic Language Models for Protein Function and Property Prediction

Curated Lists

List Inclusions