The groundwork for an AI co-scientist framework for the sequential analysis of multi-batch spatial transcriptomics data that integrates interpretable mathematical models, including optimal transport, tensor factorizations, and topological data analysis, with LLM-powered agentic interpretation is presented.
Abstract
Given the challenges of pharmaceutical development, there is increasing interest in computational approaches for identifying novel drug targets. Transcriptomics technologies enable the analysis of disease mechanisms at unprecedented resolution, while spatially resolved platforms further preserve the spatial organization of cells within tissues, allowing analysis to focus on specific regions of disease localization. To derive actionable therapeutic targets from these data, it is necessary to identify mechanisms that recur across a population rather than those observed in individual patients, which requires integrated analysis across disease conditions and donor cohorts. Here, we present the groundwork for an AI co-scientist framework for the sequential analysis of multi-batch spatial transcriptomics data that integrates interpretable mathematical models, including optimal transport, tensor factorizations, and topological data analysis, with LLM-powered agentic interpretation. Leveraging the complementary information provided by these mathematical models, the agents sequentially identify regions in which the disease is localized, molecular targets within these regions, and the direction in which each target should be modulated. Applied as a proof of concept to a public Stereo-seq study of the Alzheimer's disease (AD)-affected prefrontal cortex, we demonstrate that the agents converge on the white matter as the primary locus of the moderate-disease signal and identify a loss of myelin programme usage despite increasing oligodendrocyte abundance. The agents subsequently propose 29 candidate therapeutic targets, led by an oligodendrocyte differentiation module on which multiple mathematical perspectives converge, and flag 20 genes whose relevance to AD has not been previously reported. These results are hypothesis-generating and motivate further validation in larger, independent cohorts.
Combinatorial therapies are widely used in the treatment of acute myeloid leukemia (AML) to address disease heterogeneity, adaptive resistance, and rewired signaling and metabolic states. Yet drug prioritization remains largely guided by clinical or phenotypic evidence, while the molecular mechanisms underlying effecti...
Guo-Qing Cheng, R. Soliymani, Elham Gholizadeh et al.· bioRxiv· 0 citations
Glioblastoma (GBM) is an aggressive primary brain tumor with marked molecular and cellular heterogeneity, and current treatment provides limited long-term disease control. Single-target therapies have frequently failed because tumor subpopulations can differ in signaling dependencies, adapt under treatment pressure, an...
Yu-Chen Gao· Theoretical and Natural Scie...· 0 citations
Two deep learning-based computational protocols designed to address drug mechanisms of action and predicting drug-target interactions are presented, offering scalable, cost-effective tools for accelerating therapeutic development by bridging chemical structure, molecular interactions, and systems-level biological respo...
Y. Sun, Cheng-You Liu, Zi-Hao Jing et al.· Methods in molecular biology· 0 citations
The potential of artificial intelligence in drug discovery and development, has been explored in this review, encircling several stages from identification of target to its clinical applications, followed by drug development approaches encompassing the future of artificial intelligence in the pharmaceutical sector.
Diptanu De, Sohini Chakraborty, Satarupa Banerjee· Frontiers in Oncology· 0 citations
DrugMR is a Nextflow pipeline that runs cis-Mendelian randomisation, colocalisation, SMR-HEIDI with bulk and/or single-cell QTLs, three-trait colocalisation and a phenome-wide screen for adverse effects and repurposing, applied to schizophrenia with brain pQTLs.
G. C. Cimadevila, Marie-Joe Dib, M. Bracher-Smith et al.· bioRxiv· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.