Biology subjects

Yu, T.

Publications and source records attributed to Yu, T..

At least 73 records · Page 4Linked to original sources

bioRxiv2023

Biosynthesis of H-GDGTs linked to ocean oxygen deficiency

Archaeal membrane lipids GDGTs (glycerol dialkyl glycerol tetraethers) are biomarkers used for tracking Earths historical environmental changes. Among these GDGTs, the H-shaped GDGTs (H-GDGTs, or GMGTs) represent a less-explored and often overlooked subset, with its biosynthetic pathway and geological significance remaining elusive. Here, we identified the gene responsible for biosynthesizing H-GDGTs, which encodes to a radical S-adenosyl-L-methionine (SAM) enzyme, named as H-GDGTs bridge synthase (Hbs). Heterologous expression of the gene hbs in a methanogen, as well as in vitro activity assay using the purified Hbs enzyme were performed. Additionally, we found that the genes encoding Hbs are exclusively present in obligate anaerobic archaea genomes and the metagenomes obtained from oxygen-deficient environments, but not in oxic settings. The H-GDGTs lipids were also consistently enriched in the modern oxygen-deficient environments, and remarkably accumulated in ancient sediments during oceanic anoxic event-2 (OAE-2, [~]94 million years ago) period. Our findings indicate H-GDGTs holds significant promise as a novel biomarker for studying historical ocean oxygen deficiency supported by a well-established biological basis.

microbiology↗

bioRxiv2023

A peptide-centric local stability assay to unveil protein targets of diverse ligands

While tremendous progress has been made in chemical proteomics for identifying protein-ligand interactions, it remains challenging for proteome-wide identification of ligand-binding regions without modifying the ligands. Here, we discovered that "disruptive trypsinization" amplifies the readout of ligand-induced protein local stability shifts, and explored this notion in developing "peptide-centric local stability assay" (PELSA), a modification-free approach which achieves unprecedented sensitivity in proteome-wide target identification and binding-region determination. We demonstrate the versatility of PELSA by investigating the interactions across various biological contexts including drug-target interactions, metabolism, epitope mapping, metal proteomics, and post-translational modification recognition. A PELSA study of the oncometabolite R2HG revealed functional insights about its targets and pathogenic processes in both cancer and immune cells. Thus, beyond offering users unprecedented sensitivity for characterizing diverse target-ligand interactions, PELSA supports informative screening and hypothesis generation studies throughout life science.

biochemistry↗

bioRxiv2023

BTR: A Bioinformatics Tool Recommendation System

MotivationThe rapid expansion of Bioinformatics research has resulted in a vast array of computational tools utilized in the development of scientific analysis pipelines. However, constructing these pipelines is a laborious and intricate task, one which demands extensive domain knowledge and careful consideration at all stages. As the Bioinformatics landscape continues to evolve, researchers, both novice and expert, may find themselves overwhelmed when working in unfamiliar fields. Consequently, this may result in the selection of unsuitable or suboptimal tools during workflow development. ResultsIn this paper, we propose the Bioinformatics Tool Recommendation system (BTR), an innovative deep learning model designed to recommend the most suitable tools for a given workflow-in-progress. BTR utilizes recent advances in graph neural network technology and introduces a novel approach, representing the entire workflow as a graph to capture essential context and structural information. Additionally, natural language processing techniques are integrated to enhance the quality of tool recommendations by analyzing associated tool descriptions. Experiments demonstrate that BTR outperforms the existing Galaxy tool recommendation system, highlighting its potential to greatly facilitate scientific workflow construction. Availability and implementationThe Python source code is available at https://github.com/ryangreenj/bioinformatics_tool_recommendation

bioinformatics↗

bioRxiv2023

Expanded palette of RNA base editors for comprehensive RBP-RNA interactome studies

RNA binding proteins (RBPs) are key regulators of RNA processing and cellular function. Technologies to discover RNA targets of RBPs such as TRIBE (targets of RNA binding proteins identified by editing) and STAMP (surveying targets by APOBEC1 mediated profiling) utilize fusions of RNA base-editors (rBEs) to RBPs to circumvent the limitations of immunoprecipitation (CLIP)-based methods that require enzymatic digestion and large amounts of input material. To broaden the repertoire of rBEs suitable for editing-based RBP-RNA interaction studies, we have devised experimental and computational assays in a framework called PRINTER (protein-RNA interaction-based triaging of enzymes that edit RNA) to assess over thirty A-to-I and C-to-U rBEs, allowing us to identify rBEs that expand the characterization of binding patterns for both sequence-specific and broad-binding RBPs. We also propose specific rBEs suitable for dual-RBP applications. We show that the choice between single or multiple rBEs to fuse with a given RBP or pair of RBPs hinges on the editing biases of the rBEs and the binding preferences of the RBPs themselves. We believe our study streamlines and enhances the selection of rBEs for the next generation of RBP-RNA target discovery.

genomics↗

bioRxiv2023

Exposure to titanium dioxide nanoparticles accelerates abnormal fat deposition through the mediation of the ROS via the AGE-RAGE signaling pathway in mice

Titanium dioxide nanoparticles (TiO2 NPs) are widely added to various types of foods as food additives. Previous studies have shown that TiO2 NPs exposure can cause abnormal deposition of adipose tissue in organisms, resulting in lipid metabolism disorder. However, the potential molecular mechanisms underlying TiO2 NPs effects have yet to be elucidated. In this study, our data indicated that TiO2 NPs (100 mg/Kg BW, 20 nm) accelerated abnormal fat deposition in the epididymal adipose tissues and disturbed the level of blood glucose subsequently in normal-fat diet mice. Further studies showed that TiO2 NPs at a concentration of 100 {micro}g/mL significantly induced the proliferation and differentiation of 3T3-L1 preadipocytes. Mechanistic studies we revealed that TiO2 NPs induced the overproduction of reactive oxygen species (ROS) in the cytoplasm, which subsequently upregulated the expression of receptors for advanced glycation end products (AGEs) leading to lipid accumulation in the adipocytes with a higher level of ROS. However, the antioxidant N-acetylcysteine (NAC) was a therapeutic potential for lipid overaccumulation in the adipocytes. This study provides insight into the mechanism underlying fat deposition induced by TiO2 NPs and highlighted the need for reevaluation of food-grade TiO2 NPs exposure in daily life.

cell biology↗

bioRxiv2023

Sale of critically endangered sharks in the United States

Shark meat is widely available in the United States in grocery stores and seafood markets. The meat is often mislabeled or generically labeled as "shark". The ambiguity of these generic labels makes it challenging to assess the conservation implications of this practice and for consumers to avoid species with high mercury concentrations. For this study we purchased and DNA barcoded 30 shark products purchased in the United States to determine their species identity and conservation status. These samples consisted of 19 filets sold in grocery stores, seafood markets, and Asian specialty markets (mostly in North Carolina) and 11 ordered online as "jerky". 70% of samples were "soft mislabeled" (i.e., labeled generically as shark but not as a specific species). Of the nine samples labeled to species, eight were mislabeled (e.g., spinner shark labeled as mako shark). Only one sample was correctly labeled. All 30 samples were identified as shark and came from 11 different species, including three species listed by the IUCN as Critically Endangered: great hammerhead, scalloped hammerhead, and tope. The first two species have been found to contain very high levels of mercury, illustrating the implications of seafood mislabeling for human health. The widespread availability of shark meat in U.S. grocery stores is surprising given the dramatic decline of shark populations globally. Moreover, the fact that nearly all shark meat is either mislabeled or not labeled to species amplifies the problem. Accurate, verified product labels for shark meat would benefit consumers and shark conservation efforts, and should be a priority for the seafood industry.

genetics↗

bioRxiv2023

Aub, Vasa and Armi localization to phase separated nuage is dispensable for piRNA biogenesis and transposon silencing in Drosophila

From nematodes to placental mammals, key components of the germline transposon silencing piRNAs pathway localize to phase separated perinuclear granules. In Drosophila, the PIWI protein Aub, DEAD box protein Vasa and helicase Armi localize to nuage granules and are required for ping-pong piRNA amplification and phased piRNA processing. Drosophila piRNA mutants lead to genome instability and Chk2 kinase DNA damage signaling. By systematically analyzing piRNA pathway organization, small RNA production, and long RNA expression in single piRNA mutants and corresponding chk2/mnk double mutants, we show that Chk2 activation disrupts nuage localization of Aub and Vasa, and that the HP1 homolog Rhino, which drives piRNA precursor transcription, is required for Aub, Vasa, and Armi localization to nuage. However, these studies also show that ping-pong amplification and phased piRNA biogenesis are independent of nuage localization of Vasa, Aub and Armi. Dispersed cytoplasmic proteins thus appear to mediate these essential piRNA pathway functions.

developmental biology↗

bioRxiv2023

ProFAST: a fast and scalable factor analysis for spatially aware dimension reduction of multi-section spatial transcriptomics data

Biological techniques for spatially resolved transcriptomics (SRT) have advanced rapidly in both throughput and spatial resolution for a single spatial location. This progress necessitates the development of efficient and scalable spatial dimension reduction methods that can handle large-scale SRT data from multiple sections. Here, we developed FAST as a fast and efficient generalized probabilistic factor analysis for spatially aware dimension reduction, which simultaneously accounts for the count nature of SRT data and extracts a low-dimensional representation of SRT data across multiple sections, while preserving biological effects with consideration of spatial smoothness among nearby locations. Compared with existing methods, FAST uniquely models the count data across multiple sections while using a local spatial dependence with scalable computational complexity. Using both simulated and real datasets, we demonstrated the improved correlation between FAST estimated embeddings and annotated cell/domain types. Furthermore, FAST exhibits remarkable speed, with only FAST being applicable to analyze a mouse embryo Stereo-seq dataset with >2.3 million locations in only 2 hours. More importantly, FAST identified the differential activities of immune-related transcription factors between tumor and non-tumor clusters and also predicted a carcinogenesis factor CCNH as the upstream regulator of differentially expressed genes in a breast cancer Xenium dataset.

genomics↗

bioRxiv2023

HES1 and ID3 temporally regulate p27 levels and endothelial cell flow-mediated quiescence depth

BackgroundEndothelial cells regulate their cell cycle as blood vessels remodel and transition to quiescence downstream of blood flow-induced mechanotransduction. Laminar blood flow leads to quiescence, but how flow-mediated quiescence is established and maintained is poorly understood. MethodsPrimary human endothelial cells were exposed to laminar flow regimens and gene expression manipulations, and quiescence depth was analyzed via time to cell cycle re-entry after flow cessation. Mouse and zebrafish endothelial expression patterns were examined via scRNA seq analysis, and mutant or morphant fish lacking p27 were analyzed for endothelial cell cycle regulation and in vivo cellular behaviors. ResultsArterial flow-exposed endothelial cells had a distinct transcriptome, and they first entered a deep quiescence, then transitioned to shallow quiescence under homeostatic maintenance conditions. In contrast, venous-flow exposed endothelial cells entered deep quiescence early that did not change with homeostasis. The cell cycle inhibitor p27 (CDKN1B) was required to establish endothelial flow-mediated quiescence, and expression levels positively correlated with quiescence depth. p27 loss in vivo led to endothelial cell cycle upregulation and ectopic sprouting, consistent with loss of quiescence. HES1 and ID3, transcriptional repressors of p27 upregulated by arterial flow, were required for quiescence depth changes and the reduced p27 levels associated with shallow quiescence. ConclusionsEndothelial cell flow-mediated quiescence has unique properties and temporal regulation of quiescence depth that depends on the flow stimulus. These findings are consistent with a model whereby flow-mediated endothelial cell quiescence depth is temporally regulated downstream of p27 transcriptional regulation by HES1 and ID3. The findings are important in understanding endothelial cell quiescence mis-regulation that leads to vascular dysfunction and disease. GRAPHICAL ABSTRACT O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=197 SRC="FIGDIR/small/544403v2_ufig1.gif" ALT="Figure 1"> View larger version (43K): org.highwire.dtl.DTLVardef@1137a1forg.highwire.dtl.DTLVardef@1b6bf1dorg.highwire.dtl.DTLVardef@bd4965org.highwire.dtl.DTLVardef@11dbb9b_HPS_FORMAT_FIGEXP M_FIG C_FIG HIGHLIGHTSO_LIDifferent quiescence stimuli lead to distinct transcriptional and functional quiescence profiles in endothelial cells C_LIO_LIp27 is required for endothelial cell quiescence and depth is temporally regulated in a flow stimulus-dependent manner that correlates with p27 levels and flow-regulated repressors HES1 and ID3 C_LIO_LIp27 is expressed in endothelial cells according to flow magnitude in vivo and is functionally required for cell cycle regulation and sprouting in vivo C_LI

cell biology↗

bioRxiv2023

Global Marine Cold Seep Metagenomes Reveal Diversity of Taxonomy, Metabolic Function, and Natural Products

Cold seeps in the deep sea are closely linked to energy exploration as well as global climate change. The alkane-dominated chemical energy-driven model makes cold seeps an oasis of deep-sea life, showcasing an unparalleled reservoir of microbial genetic diversity. By analyzing 113 metagenomes collected from 14 global sites across 5 cold seep types, we present a comprehensive Cold Seep Microbiomic Database (CSMD) to archive the genomic and functional diversity of cold seep microbiome. The CSMD includes over 49 million non-redundant genes and 3175 metagenome-assembled genomes (MAGs), which represent 1897 species spanning 106 phyla. In addition, beta diversity analysis indicates that both sampling site and cold seep type have substantial impact on the prokaryotic microbiome community composition. Heterotrophic and anaerobic metabolisms are prevalent in microbial communities, accompanied by considerable mixotrophs and facultative anaerobes, indicating the versatile metabolic potential in cold seeps. Furthermore, secondary metabolic gene cluster analysis indicates that at least 98.81% of the sequences encode potentially novel natural products. These natural products are dominated by ribosomal processing peptides, which are widely distributed in archaea and bacteria. Overall, the CSMD represents a valuable resource which would enhance the understanding and utilization of global cold seep microbiomes.

microbiology↗

bioRxiv2023

Endothelial Cell SMAD6 Balances ACVRL1/Alk1 Function to Regulate Adherens Junctions and Hepatic Vascular Development

BMP signaling is critical to blood vessel formation and function, but how pathway components regulate vascular development is not well-understood. Here we find that inhibitory SMAD6 functions in endothelial cells to negatively regulate ALK1/ACVRL1-mediated responses, and it is required to prevent vessel dysmorphogenesis and hemorrhage in the embryonic liver vasculature. Reduced Alk1 gene dosage rescued embryonic hepatic hemorrhage and microvascular capillarization induced by Smad6 deletion in endothelial cells in vivo. At the cellular level, co-depletion of Smad6 and Alk1 rescued the destabilized junctions and impaired barrier function of endothelial cells depleted for SMAD6 alone. At the mechanistic level, blockade of actomyosin contractility or increased PI3K signaling rescued endothelial junction defects induced by SMAD6 loss. Thus, SMAD6 normally modulates ALK1 function in endothelial cells to regulate PI3K signaling and contractility, and SMAD6 loss increases signaling through ALK1 that disrupts endothelial junctions. ALK1 loss-of-function also disrupts vascular development and function, indicating that balanced ALK1 signaling is crucial for proper vascular development and identifying ALK1 as a "Goldilocks" pathway in vascular biology regulated by SMAD6.

developmental biology↗

bioRxiv2023

Larger and denser: an optimal design for surface grids of EMG electrodes to identify greater and more representative samples of motor units

The spinal motor neurons are the only neural cells whose individual activity can be non-invasively identified. This is usually done using grids of surface electromyographic (EMG) electrodes and source separation algorithms; an approach called EMG decomposition. In this study, we combined computational and experimental analyses to assess how the design parameters of grids of electrodes influence the number and the properties of the identified motor units. We first computed the percentage of motor units that could be theoretically discriminated within a pool of 200 simulated motor units when decomposing EMG signals recorded with grids of various sizes and interelectrode distances (IED). Increasing the density, the number of electrodes, and the size of the grids, increased the number of motor units that our decomposition algorithm could theoretically discriminate, i.e., up to 83.5% of the simulated pool (range across conditions: 30.5-83.5%). We then identified motor units from experimental EMG signals recorded in six participants with grids of various sizes (range: 2-36 cm2) and IED (range: 4-16 mm). The configuration with the largest number of electrodes and the shortest IED maximized the number of identified motor units (56{+/-}14; range: 39-79) and the percentage of early recruited motor units within these samples (29{+/-}14%). Finally, the number of identified motor units further increased with a prototyped grid of 256 electrodes and an IED of 2 mm. Taken together, our results showed that larger and denser surface grids of electrodes allow to identify a more representative pool of motor units than currently reported in experimental studies. Significance statementThe application of source separation methods to multi-channel EMG signals recorded with grids of electrodes enables users to accurately identify the activity of individual motor units. However, the design parameters of these grids have never been discussed. They are usually arbitrarily fixed, often based on commercial availability. Here, we showed that using larger and denser grids of electrodes than conventionally proposed drastically increases the number of identified motor units. The samples of identified units are more balanced between early- and late-recruited motor units. Thus, these grids provide a more representative sampling of the active motor unit population. Gathering large datasets of motor units using large and dense grids will impact the study of motor control, neuromuscular modelling, and human-machine interfacing.

neuroscience↗

bioRxiv2023

Self-supervised learning for DNA sequences with circular dilated convolutional networks

Due to their intrinsic properties, DNA molecules commonly exhibit long-range interactions along a linear sequence representation. Taking this information into account when modeling DNA sequences is therefore important for obtaining more accurate sequence-based inference. Many deep learning methods have recently been developed for this purpose, but they still suffer from two major issues. First, the existing methods can only handle short DNA fragments, thereby losing longerrange interactions. Second, the current methods require massive supervised labeling while missing most order information within the sequences. Consequently, there is a need to develop an efficient deep neural network modeling framework to extract wide contextual information for more accurate sequence-based inference tasks. Our new framework, named Revolution, takes full DNA sequences as input, without any condensation, and can give accurate predictions for DNA sequences up to 10kbp. In variant effect prediction, our method increases the Area Under the Receiver Operating Characteristics (AUROC) by 19.61% on 49 human tissues on average. Revolution is also demonstrated to work on the plant sequences by improving 2.36% AUROC on average for predicting open chromatin regions (OCRs). The data, models, and code can be freely accessed at https://github.com/wiedersehne/Revolution-DNAPretraining.

bioinformatics↗

bioRxiv2023

G3DC: a Gene-Graph-Guided selective Deep Clustering method for single cell RNA-seq data

Single-Cell RNA sequencing (scRNA-seq) technology measures the expression of thousands of genes at the cellular level. Analyzing single cell transcriptome allows the identification of heterogeneous cell groups, cellular-level regulations, and the trajectory of cell development. An important aspect in the analyses of scRNA-seq data is the clustering of cells, which is hampered by issues such as high dimensionality, cell type imbalance, redundancy, and dropout. Given cells of each type are functionally consistent, incorporating biological relations between genes may improve the clustering results. Here, we develop a deep embedded clustering method, G3DC, that incorporates a graph loss based on existing gene network, together with a reconstruction loss to achieve both discriminative and informative embedding. The involvement of the gene network strengthens clustering performance, while helping the selection of functionally coherent genes that contribute to the clustering results. In addition, this method is well adapted to the sparse and zero-inflated scRNA-seq data with the{ell} 2,1-norm involved. Extensive experiments have shown that G3DC offers high clustering accuracy with regard to agreement with true cell types, outperforming other leading single-cell clustering methods. In addition, G3DC selects biologically relevant genes that contribute to the clustering, providing insight into biological functionality that differentiate cell groups.

bioinformatics↗

bioRxiv2023

Systematic elucidation of genetic mechanisms underlying cholesterol uptake

Genetic variation contributes greatly to LDL cholesterol (LDL-C) levels and coronary artery disease risk. By combining analysis of rare coding variants from the UK Biobank and genome-scale CRISPR-Cas9 knockout and activation screening, we have substantially improved the identification of genes whose disruption alters serum LDL-C levels. We identify 21 genes in which rare coding variants significantly alter LDL-C levels at least partially through altered LDL-C uptake. We use co-essentiality-based gene module analysis to show that dysfunction of the RAB10 vesicle transport pathway leads to hypercholesterolemia in humans and mice by impairing surface LDL receptor levels. Further, we demonstrate that loss of function of OTX2 leads to robust reduction in serum LDL-C levels in mice and humans by increasing cellular LDL-C uptake. Altogether, we present an integrated approach that improves our understanding of genetic regulators of LDL-C levels and provides a roadmap for further efforts to dissect complex human disease genetics.

genetics↗

bioRxiv2023

MitoQuicLy: a high-throughput method for quantifying cell-free DNA from human plasma, serum, and saliva

Circulating cell-free mitochondrial DNA (cf-mtDNA) is an emerging biomarker of psychobiological stress and disease which predicts mortality and is associated with various disease states. To evaluate the contribution of cf-mtDNA to health and disease states, standardized high-throughput procedures are needed to quantify cf-mtDNA in relevant biofluids. Here, we describe MitoQuicLy: Mitochondrial DNA Quantification in cell-free samples by Lysis. We demonstrate high agreement between MitoQuicLy and the commonly used column-based method, although MitoQuicLy is faster, cheaper, and requires a smaller input sample volume. Using 10 {micro}L of input volume with MitoQuicLy, we quantify cf-mtDNA levels from three commonly used plasma tube types, two serum tube types, and saliva. We detect, as expected, significant inter-individual differences in cf-mtDNA across different biofluids. However, cf-mtDNA levels between concurrently collected plasma, serum, and saliva from the same individual differ on average by up to two orders of magnitude and are poorly correlated with one another, pointing to different cf-mtDNA biology or regulation between commonly used biofluids in clinical and research settings. Moreover, in a small sample of healthy women and men (n=34), we show that blood and saliva cf-mtDNAs correlate with clinical biomarkers differently depending on the sample used. The biological divergences revealed between biofluids, together with the lysis-based, cost-effective, and scalable MitoQuicLy protocol for biofluid cf-mtDNA quantification, provide a foundation to examine the biological origin and significance of cf-mtDNA to human health.

molecular biology↗

bioRxiv2022

Experimental study of modified human tendon stem cells to promote graft ligamentation and tendon-bone healing after anterior cruciate ligament reconstruction

Restoring the normal structure and function of injured tendons is one of the biggest challenges faced by the Department of Orthopedics and Sports Medicine. Tendon-derived stem cells (TDSCs), a new type of pluripotent stem cells with multidirectional differentiation potential, are expected to be promising cell seeds for the treatment of tendon injury and tendon-bone healing in the future. In this study, tendon stem cells were successfully isolated from human tissues, which were positive for markers CD44, CD90, and CD105, and exhibited clonality and multilineage differentiation ability. Analysis of single-cell sequencing results and mass spectrometry identification results showed that there were differences in protein expression during CTGF-induced TDSC tendon differentiation. Reverse Co-IP, qPCR, WB, and immunofluorescence detection all confirmed that CTGF directly interacts with KIT, thereby mediating the transcription factor HES1 to regulate the Wnt/{beta}-catenin signaling pathway (GSK3{beta}, {beta}-catenin, TCF4). ChIP-qPCR and dual-luciferase reporter gene assays indicated that HES1 regulates stem cell differentiation by directly regulating the expression of GSK3{beta} in the Wnt/{beta}-catenin pathway. Rats were treated with TDSCs overexpressing the KIT gene after repair surgery. This method had a more ideal recovery effect than other methods through animal behavioral scores, mechanical properties testing, and HE staining tissue observation. This study found that the use of modified human tendon stem cells (hTDSC) could promote graft ligamentization and tendon-bone healing after ACL reconstruction, which could provide an effective way for faster and better recovery from tendon injury.

cell biology↗

bioRxiv2022

A Community Challenge to Predict Clinical Outcomes After Immune Checkpoint Blockade in Non-Small Cell Lung Cancer

PurposePredictive biomarkers of immune checkpoint inhibitors (ICIs) efficacy are currently lacking for non-small cell lung cancer (NSCLC). Here, we describe the results from the Anti-PD-1 Response Prediction DREAM Challenge, a crowdsourced initiative that enabled the assessment of predictive models by using data from two randomized controlled clinical trials (RCTs) of ICIs in first-line metastatic NSCLC. MethodsParticipants developed and trained models using public resources. These were evaluated with data from the CheckMate 026 trial (NCT02041533), according to the model-to-data paradigm to maintain patient confidentiality. The generalizability of the models with the best predictive performance was assessed using data from the CheckMate 227 trial (NCT02477826). Both trials were phase III RCTs with a chemotherapy control arm, which supported the differentiation between predictive and prognostic models. Isolated model containers were evaluated using a bespoke strategy that considered the challenges of handling transcriptome data from clinical trials. ResultsA total of 59 teams participated, with 417 models submitted. Multiple predictive models, as opposed to a prognostic model, were generated for predicting overall survival, progression-free survival, and progressive disease status with ICIs. Variables within the models submitted by participants included tumor mutational burden (TMB), programmed death ligand 1 (PD-L1) expression, and gene-expression-based signatures. The bestperforming models showed improved predictive power over reference variables, including TMB or PD-L1. ConclusionThis DREAM Challenge is the first successful attempt to use protected phase III clinical data for a crowdsourced effort towards generating predictive models for ICIs clinical outcomes and could serve as a blueprint for similar efforts in other tumor types and disease states, setting a benchmark for future studies aiming to identify biomarkers predictive of ICIs efficacy. Context summaryO_ST_ABSKey objectiveC_ST_ABSNot all patients with non-small cell lung cancer (NSCLC) eligible for immune checkpoint inhibitor (ICIs) respond to treatment, but accurate predictive biomarkers of ICIs clinical outcomes are currently lacking. This crowdsourced initiative enabled the robust assessment of predictive models using data from two randomized clinical trials of first-line ICI in metastatic NSCLC. Knowledge generatedModels submitted indicate that a combination of programmed death ligand 1 (PD-L1), tumor mutational burden (TMB), and immune gene signatures might be able to identify patients more likely to respond to ICIs. TMB and PD-L1 seemed important to predict progression-free survival and overall survival. Mechanisms including apoptosis, T-cell crosstalk, and adaptive immune resistance appeared essential to predict response. Relevance

cancer biology↗