Lifting Biomolecular Data Acquisition

2025-12-17

AI Summary

Wet-lab experiments suffer from low information density, hindering large-scale ML-driven biological discovery. Method: This paper introduces a neural compressed sensing framework that enables parallel measurement and end-to-end differentiable deconvolution of molecular activities within the molecular activity function space. It pioneers a “wet-experiment–algorithm” co-design paradigm, integrating neural-network-driven compressed sampling, functional-space representation, and differentiable experimental decoupling—extending compressed sensing to non-Euclidean function-space modeling for the first time. Contribution/Results: Theoretical analysis proves a 10- to 100-fold (1–2 orders of magnitude) increase in information density. Empirical validation in antibody screening and cell therapy demonstrates a 10×–100× gain in information yield per experiment, significantly accelerating high-throughput biological molecule discovery cycles.

Technology Category

• Machine Learning: Active Learning
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Search and Optimization: Mixed Discrete/Continuous Search

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Web Mining and Content Analysis: Machine learning and data science for the Web
• Security and Privacy: Large-scale security measurements

Abstract

One strategy to scale up ML-driven science is to increase wet lab experiments' information density. We present a method based on a neural extension of compressed sensing to function space. We measure the activity of multiple different molecules simultaneously, rather than individually. Then, we deconvolute the molecule-activity map during model training. Co-design of wet lab experiments and learning algorithms provably leads to orders-of-magnitude gains in information density. We demonstrate on antibodies and cell therapies.

Problem

• Increase information density in wet lab experiments
• Simultaneously measure multiple molecular activities
• Deconvolute molecule-activity maps during model training

Innovation

• Neural compressed sensing for function space
• Simultaneous measurement of multiple molecules
• Co-design of wet lab and learning algorithms