Integrating benchwork with generative AI development: deciphering cell size regulation while driving predictive AI

Artificial intelligence promises to dramatically accelerate preclinical biomedical research, yet ensuring it provides accurate and relevant information remains a challenge. Current approaches may fabricate references and often lack the context needed to fully address scientific questions. Consequently, AI has not been widely adopted for applications such as hit selection from biological screens or experiment design.
To address these challenges, BenchSci, a leader in AI-based solutions for preclinical research, is partnering with Dr. Moffat’s lab, renowned for its genome-wide genetic screens in human cell lines. In this collaboration, a postdoctoral fellow will test an advanced question-answering system under development at BenchSci. This tool aims to assist in hit selection, rapidly identifying relevant hits for understanding disease progression. It will also be used in ongoing studies on proteins responding to salt stress, exploring their links to certain cancers and metabolic diseases.
By collaborating directly with Dr. Moffat’s research group, BenchSci can quickly optimize the prototype for generating hypotheses and planning new experiments. As the postdoctoral fellow conducts research, the resulting data will help BenchSci develop more predictive systems, potentially enabling AI to forecast experimental outcomes. Moreover, the system will evolve to suggest subsequent experimental steps, enhancing its role as an effective research assistant.
This partnership aims to accelerate the development of cutting-edge AI systems for biologists and advance critical genetic interaction studies of disease pathways.

Faculty Supervisor:

Jason Moffat

Student:

Partner:

BenchSci

Discipline:

Life Sciences

Sector:

Professional, scientific and technical services

University:

University of Toronto

Program: