Epstein, A. | Search bioRxiv

bioRxiv2023

Clustering Protein Binding Pockets and Identifying Potential Drug Interactions: A Novel Ligand-based Featurization Method

Protein-ligand interactions are essential to drug discovery and drug development efforts. Desirable on-target or multi-target interactions are a first step in finding an effective therapeutic; undesirable off-target interactions are a first step in assessing safety. In this work, we introduce a novel ligand-based featurization and mapping of human protein pockets to identify closely related protein targets, and to project novel drugs into a hybrid protein-ligand feature space to identify their likely protein interactions. Using structure-based template matches from PDB, protein pockets are featurized by the ligands which bind to their best co-complex template matches. The simplicity and interpretability of this approach provides a granular characterization of the human proteome at the protein pocket level instead of the traditional protein-level characterization by family, function, or pathway. We demonstrate the power of this featurization method by clustering a subset of the human proteome and evaluating the predicted cluster associations of over 7,000 compounds.

synthetic biology↗

bioRxiv2022

A global view of aging and Alzheimer's pathogenesis-associated cell population dynamics and molecular signatures in the human and mouse brains

Conventional single-cell genomics approaches are limited by throughput and thus may have failed to capture aspects of the molecular signatures and dynamics of rare cell types associated with aging and diseases. Here, we developed EasySci, an extensively improved single-cell combinatorial indexing strategy, for investigating the age-dependent dynamics of transcription and chromatin accessibility across diverse brain cell types. We profiled [~]1.5 million single-cell transcriptomes and [~]400,000 single-cell chromatin accessibility profiles across mouse brains spanning different ages, genotypes, and both sexes. With a novel computational framework designed for characterizing cellular subtypes based on the expression of both genes and exons, we identified > 300 cell subtypes and deciphered their underlying molecular programs and spatial locations especially for rare cell types (e.g., pinealocytes, tanycytes). Leveraging these data, we generated a global readout of age-dependent changes at cell subtype resolution, providing insights into cell types that expand (e.g., rare astrocytes and vascular leptomeningeal cells in the olfactory bulb, reactive microglia, and oligodendrocytes) or are depleted (e.g., neuronal progenitors, neuroblasts, committed oligodendrocyte precursors) as age progresses. Furthermore, we explored cell-type-specific responses to genetic perturbations associated with Alzheimers disease (AD) and identified rare cell types depleted (e.g., mt-Cytb+, mt-Rnr2+ choroid plexus epithelial cells) or enriched (e.g., Col25a1+, Ndrg1+ interbrain and midbrain neurons) in both AD models. Key findings are consistent between males and females, validated across the transcriptome, chromatin accessibility, and spatial analyses. Finally, we profiled a total of 118,240 single-nuclei transcriptomes from twenty-four post-mortem human brain samples derived from control and AD patients, revealing highly cell-type-specific and region-specific gene expression changes associated with AD pathogenesis. Critical AD-associated gene signatures were validated in both human and mice. In summary, these data comprise a rich resource for exploring cell-type-specific dynamics and the underlying molecular mechanisms in normal and pathological mammalian aging.

genomics↗