Logo
Umbilical Life

Computational Imaging Scientist – AI and Toxicology Prediction

Umbilical Life, San Francisco, California, United States, 94199


Are you an Imaging Expert? Want to help revolutionalise Drug Discovery? Want to be part of the AI revolution?

Job Title : Computational Imaging Scientist – AI and Toxicology Prediction

Location : San Francisco, Bay Area

Company Overview :

I am working with a cutting-edge company revolutionizing toxicology through the power of AI and computational imaging.

Their mission is to

replace traditional chemical toxicity testing with AI-driven predictive model s that enhance safety in the pharmaceutical, food, cosmetics, and agricultural sectors.

Leveraging a unique combination of

high-content microscopy, chemical data, and advanced neural network architectures , we are building pioneering models that

translate chemical structures

into biological responses, offering an innovative, scalable approach to

toxicity prediction .

Position Summary :

Looking to ID an expert in

computational imaging

with a

strong background in AI and machine learning

to join our team.

In this role, you will spearhead the development of

high-content imaging pipelines , integrating

primary cell microscopy data

with

neural network-based model s to generate insights into

chemical toxicity.

This is a unique opportunity to contribute to a transformative field, where your expertise will be pivotal in creating AI models capable of rivaling and surpassing traditional physical experiments.

Key Responsibilities :

Microscopy and Image Data Generation :

Develop and oversee high-content imaging workflows, generating extensive microscopy datasets on primary cells.

Optimize imaging modalities for accuracy and detail, facilitating downstream integration with chemical and biochemical data.

Lead initiatives to generate over 40 million cell images across a diverse set of 50,000 small molecules at various concentration levels.

Machine Learning Model Development :

Design, train, and refine computer vision models, including Convolutional Neural Networks (CNNs) and Vision Transformers, to classify images, detect phenotypic features, and perform anomaly detection.

Develop and apply image and text encoder models to enable text-to-image predictions, enhancing the ability to predict chemical-biology interactions.

Innovate in multi-modal model development to create a "chemical-to-biology" predictive model, mapping chemical structures to biological phenotypes.

High-Throughput Phenotypic Analysis :

Develop algorithms to detect and quantify phenotypes in high-content imaging, capturing the "phenotypic landscape" and identifying biologically diverse responses.

Build and validate machine learning models that utilize high-resolution imaging data to accurately predict human toxicity profiles.

AI and Neural Network Research :

Conduct research on novel neural network architectures for imaging applications, focusing on deep learning models, transformer models, and large-scale pretraining.

Collaborate with the data science and AI teams to integrate imaging data into large language models (LLMs) and multi-modal representations for in-depth data exploration.

Collaboration and Innovation :

Work cross-functionally with chemists, biologists, and toxicologists to align imaging processes with broader project goals.

Actively contribute to a collaborative research environment, presenting findings and publishing research to advance the field of AI in toxicology.

Qualifications :

Education : PhD or Master’s degree in Computational Imaging, Computer Vision, Biomedical Engineering, or a related field.

Experience :

You would consider yourself an expert

in imaging science, with a strong background in high-content microscopy, cell imaging, and computational biology.

Technical Expertise :

Proficiency in

Python

and

relevant machine learning libraries

(e.g., PyTorch, TensorFlow).

Extensive experience with Convolutional Neural Networks ( CNNs ),

Vision Transformers , and other

deep learning architectures

for imaging modalities.

Background in

multi-modal model

development,

integrating imaging data

with

structured data

(e.g., chemical or biochemical datasets).

Expertise in anomaly detection, image classification, and other computer vision techniques.

Domain Knowledge : Familiarity with chemical and biological data and experience applying imaging techniques to biochemical and toxicological contexts.

Preferred Skills :

Experience working with large-scale imaging datasets and high-throughput imaging systems.

Knowledge of phenotypic profiling, feature extraction, and image-to-structure prediction models.

Demonstrated contributions to AI-driven scientific research, including published papers and/or patents in the field.

What We Offer :

Competitive salary and benefits package.

Opportunity to work at the forefront of AI and imaging science with a team of world-class experts.

Collaborative work environment dedicated to pushing the boundaries of scientific discovery.

#J-18808-Ljbffr