Organoid Modeling of the Tumor Immune Microenvironment.

In vitro cancer cultures, including three-dimensional organoids, typically contain exclusively neoplastic epithelium but require artificial reconstitution to recapitulate the tumor microenvironment (TME). The co-culture of primary tumor epithelia with endogenous, syngeneic tumor-infiltrating lymphocytes (TILs) as a cohesive unit has been particularly elusive. Here, an air-liquid interface (ALI) method propagated patient-derived organoids (PDOs) from >100 human biopsies or mouse tumors in syngeneic immunocompetent hosts as tumor epithelia with native embedded immune cells (T, B, NK, macrophages). Robust droplet-based, single-cell simultaneous determination of gene expression and immune repertoire indicated that PDO TILs accurately preserved the original tumor T cell receptor (TCR) spectrum. Crucially, human and murine PDOs successfully modeled immune checkpoint blockade (ICB) with anti-PD-1- and/or anti-PD-L1 expanding and activating tumor antigen-specific TILs and eliciting tumor cytotoxicity. Organoid-based propagation of primary tumor epithelium en bloc with endogenous immune stroma should enable immuno-oncology investigations within the TME and facilitate personalized immunotherapy testing.

Evaluation of large language models for discovery of gene set function.

Gene set analysis is a mainstay of functional genomics, but it relies on curated databases of gene functions that are incomplete. Here we evaluate five Large Language Models (LLMs) for their ability to discover the common biological functions represented by a gene set, substantiated by supporting rationale, citations and a confidence assessment. Benchmarking against canonical gene sets from the Gene Ontology, GPT-4 confidently recovered the curated name or a more general concept (73% of cases), while benchmarking against random gene sets correctly yielded zero confidence. Gemini-Pro and Mixtral-Instruct showed ability in naming but were falsely confident for random sets, whereas Llama2-70b had poor performance overall. In gene sets derived from 'omics data, GPT-4 identified novel functions not reported by classical functional enrichment (32% of cases), which independent review indicated were largely verifiable and not hallucinations. The ability to rapidly synthesize common gene functions positions LLMs as valuable 'omics assistants.

A foundational atlas of autism protein interactions reveals molecular convergence.

Translating high-confidence (hc) autism spectrum disorder (ASD) genes into viable treatment targets remains elusive. We constructed a foundational protein-protein interaction (PPI) network in HEK293T cells involving 100 hcASD risk genes, revealing over 1,800 PPIs (87% novel). Interactors, expressed in the human brain and enriched for ASD but not schizophrenia genetic risk, converged on protein complexes involved in neurogenesis, tubulin biology, transcriptional regulation, and chromatin modification. A PPI map of 54 patient-derived missense variants identified differential physical interactions, and we leveraged AlphaFold-Multimer predictions to prioritize direct PPIs and specific variants for interrogation in Xenopus tropicalis and human forebrain organoids. A mutation in the transcription factor FOXP1 led to reconfiguration of DNA binding sites and altered development of deep cortical layer neurons in forebrain organoids. This work offers new insights into molecular mechanisms underlying ASD and describes a powerful platform to develop and test therapeutic strategies for many genetically-defined conditions.

Evaluation of large language models for discovery of gene set function.

Gene set analysis is a mainstay of functional genomics, but it relies on manually curated databases of gene functions that are incomplete and unaware of biological context. Here we evaluate the ability of OpenAI's GPT-4, a Large Language Model (LLM), to develop hypotheses about common gene functions from its embedded biomedical knowledge. We created a GPT-4 pipeline to label gene sets with names that summarize their consensus functions, substantiated by analysis text and citations. Benchmarking against named gene sets in the Gene Ontology, GPT-4 generated very similar names in 50% of cases, while in most remaining cases it recovered the name of a more general concept. In gene sets discovered in 'omics data, GPT-4 names were more informative than gene set enrichment, with supporting statements and citations that largely verified in human review. The ability to rapidly synthesize common gene functions positions LLMs as valuable functional genomics assistants.

Acute kidney injury detection using refined and physiological-feature augmented urine output.

Acute kidney injury (AKI) is common in the intensive care unit, where it is associated with increased mortality. AKI is often defined using creatinine and urine output criteria. The creatinine-based definition is more reliable but less expedient, whereas the urine output based definition is rapid but less reliable. Our goal is to examine the urine output criterion and augment it with physiological features for better agreement with creatinine-based definitions of AKI. The objectives are threefold: (1) to characterize the baseline agreement of urine output and creatinine definitions of AKI; (2) to refine the urine output criteria to identify the thresholds that best agree with the creatinine-based definition; and (3) to build generalized estimating equation (GEE) and generalized linear mixed-effects (GLME) models with static and time-varying features to improve the accuracy of a near-real-time marker for AKI. We performed a retrospective observational study using data from two independent critical care databases, MIMIC-III and eICU, for critically ill patients who developed AKI in intensive care units. We found that the conventional urine output criterion (6 hr, 0.5 ml/kg/h) has specificity and sensitivity of 0.49 and 0.54 for MIMIC-III database; and specificity and sensitivity of 0.38 and 0.56 for eICU. Secondly, urine output thresholds of 12 hours and 0.6 ml/kg/h have specificity and sensitivity of 0.58 and 0.48 for MIMIC-III; and urine output thresholds of 10 hours and 0.6 ml/kg/h have specificity and sensitivity of 0.49 and 0.48 for eICU. Thirdly, the GEE model of four hours duration augmented with static and time-varying features can achieve a specificity and sensitivity of 0.66 and 0.61 for MIMIC-III; and specificity and sensitivity of 0.66 and 0.64 for eICU. The GLME model of four hours duration augmented with static and time-varying features can achieve a specificity and sensitivity of 0.71 and 0.55 for MIMIC-III; and specificity and sensitivity of 0.66 and 0.60 for eICU. The GEE model has greater performance than the GLME model, however, the GLME model is more reflective of the variables as fixed effects or random effects. The significant improvement in performance, relative to current definitions, when augmenting with patient features, suggest the need of incorporating these features when detecting disease onset and modeling at window-level rather than patient-level.