arXiv subjects

George M. Church

Publications and source records attributed to George M. Church.

7 recordsLinked to original sources

arXiv2025

Graph AI generates neurological hypotheses validated in molecular, organoid, and clinical systems

Neurological diseases are the leading global cause of disability, yet most lack disease-modifying treatments. We present PROTON, a heterogeneous graph transformer that generates testable hypotheses across molecular, organoid, and clinical systems. To evaluate PROTON, we apply it to Parkinson's disease (PD), bipolar disorder (BD), and Alzheimer's disease (AD). In PD, PROTON linked genetic risk loci to genes essential for dopaminergic neuron survival and predicted pesticides toxic to patient-derived neurons, including the insecticide endosulfan, which ranked within the top 1.29% of predictions. In silico screens performed by PROTON reproduced six genome-wide $α$-synuclein experiments, including a split-ubiquitin yeast two-hybrid system (normalized enrichment score [NES] = 2.30, FDR-adjusted $p < 1 \times 10^{-4}$), an ascorbate peroxidase proximity labeling assay (NES = 2.16, FDR $< 1 \times 10^{-4}$), and a high-depth targeted exome sequencing study in 496 synucleinopathy patients (NES = 2.13, FDR $< 1 \times 10^{-4}$). In BD, PROTON predicted calcitriol as a candidate drug that reversed proteomic alterations observed in cortical organoids derived from BD patients. In AD, we evaluated PROTON predictions in health records from $n = 610,524$ patients at Mass General Brigham, confirming that five PROTON-predicted drugs were associated with reduced seven-year dementia risk (minimum hazard ratio = 0.63, 95% CI: 0.53-0.75, $p < 1 \times 10^{-7}$). PROTON generated neurological hypotheses that were evaluated across molecular, organoid, and clinical systems, defining a path for AI-driven discovery in neurological disease.

q-bio.QM↗

arXiv2018

Variational auto-encoding of protein sequences

Proteins are responsible for the most diverse set of functions in biology. The ability to extract information from protein sequences and to predict the effects of mutations is extremely valuable in many domains of biology and medicine. However the mapping between protein sequence and function is complex and poorly understood. Here we present an embedding of natural protein sequences using a Variational Auto-Encoder and use it to predict how mutations affect protein function. We use this unsupervised approach to cluster natural variants and learn interactions between sets of positions within a protein. This approach generally performs better than baseline methods that consider no interactions within sequences, and in some cases better than the state-of-the-art approaches that use the inverse-Potts model. This generative model can be used to computationally guide exploration of protein sequence space and to better inform rational and automatic protein design.

q-bio.QM↗

arXiv2015

New class of biological detectors for WIMPs

Weakly Interacting Massive Particles (WIMPs) may constitute a large fraction of the matter in the Universe. There are excess events in the data of DAMA/LIBRA, CoGeNT, CRESST-II, and recently CDMS-Si, which could be consistent with WIMP masses of approximately 10 GeV/c2. However, for MDM > 10 GeV/c2 null results of the CDMS-Ge, XENON, and LUX detectors may be in tension with the potential detections for certain dark matter scenarios and assuming a certain light response. We propose the use of a new class of biological dark matter (DM) detectors to further examine this light dark matter hypothesis, taking advantage of new signatures with low atomic number targets, Two types of biological DM detectors are discussed here: DNA-based detectors and enzymatic reactions (ER) based detectors. In the case of DNA-based detectors, we discuss a new implementation. In the case of ER detectors, there are four crucial phases of the detection process: a) change of state due to energy deposited by a particle; b) amplification due to the release of energy derived from the action of an enzyme on its substrate; c) sustainable but non-explosive enzymatic reaction; d) self-termination due to the denaturation of the enzyme, when the temperature is raised. This paper provides information of how to design as well as optimize these four processes.

astro-ph.IM↗

arXiv2015

Puzzle Imaging: Using Large-scale Dimensionality Reduction Algorithms for Localization

Current high-resolution imaging techniques require an intact sample that preserves spatial relationships. We here present a novel approach, "puzzle imaging," that allows imaging a spatially scrambled sample. This technique takes many spatially disordered samples, and then pieces them back together using local properties embedded within the sample. We show that puzzle imaging can efficiently produce high-resolution images using dimensionality reduction algorithms. We demonstrate the theoretical capabilities of puzzle imaging in three biological scenarios, showing that (1) relatively precise 3-dimensional brain imaging is possible; (2) the physical structure of a neural network can often be recovered based only on the neural connectivity matrix; and (3) a chemical map could be reproduced using bacteria with chemosensitive DNA and conjugative transfer. The ability to reconstruct scrambled images promises to enable imaging based on DNA sequencing of homogenized tissue samples.

q-bio.NC↗

arXiv2014

Spatial Information in Large-Scale Neural Recordings

A central issue in neural recording is that of distinguishing the activities of many neurons. Here, we develop a framework, based on Fisher information, to quantify how separable a neuron's activity is from the activities of nearby neurons. We (1) apply this framework to model information flow and spatial distinguishability for several electrical and optical neural recording methods, (2) provide analytic expressions for information content, and (3) demonstrate potential applications of the approach. This method generalizes to many recording devices that resolve objects in space and thus may be useful in the design of next-generation scalable neural recording systems.

q-bio.NC↗

arXiv2013

Physical Principles for Scalable Neural Recording

Simultaneously measuring the activities of all neurons in a mammalian brain at millisecond resolution is a challenge beyond the limits of existing techniques in neuroscience. Entirely new approaches may be required, motivating an analysis of the fundamental physical constraints on the problem. We outline the physical principles governing brain activity mapping using optical, electrical,magnetic resonance, and molecular modalities of neural recording. Focusing on the mouse brain, we analyze the scalability of each method, concentrating on the limitations imposed by spatiotemporal resolution, energy dissipation, and volume displacement. We also study the physics of powering and communicating with microscale devices embedded in brain tissue.

q-bio.NC↗

arXiv2001

A Dynamic Programming Approach to De Novo Peptide Sequencing via Tandem Mass Spectrometry

The tandem mass spectrometry fragments a large number of molecules of the same peptide sequence into charged prefix and suffix subsequences, and then measures mass/charge ratios of these ions. The de novo peptide sequencing problem is to reconstruct the peptide sequence from a given tandem mass spectral data of k ions. By implicitly transforming the spectral data into an NC-spectrum graph G=(V,E) where |V|=2k+2, we can solve this problem in O(|V|+|E|) time and O(|V|) space using dynamic programming. Our approach can be further used to discover a modified amino acid in O(|V||E|) time and to analyze data with other types of noise in O(|V||E|) time. Our algorithms have been implemented and tested on actual experimental data.

cs.CE↗