Gene-gene Relationships Research Articles

Imputing spatial transcriptomics through gene network constructed from protein language model.

Image-based spatial transcriptomic sequencing technologies have enabled the measurement of gene expression at single-cell resolution, but with a limited number of genes. Current computational approaches attempt to overcome these limitations by imputing missing genes, but face challenges regarding prediction accuracy and identification of cell populations due to the neglect of gene-gene relationships. In this context, we present stImpute, a method to impute spatial transcriptomics according to reference scRNA-seq data based on the gene network constructed from the protein language model ESM-2. Specifically, stImpute employs an autoencoder to create gene expression embeddings for both spatial transcriptomics and scRNA-seq data, which are used to identify the nearest neighboring cells between scRNA-seq and spatial transcriptomics datasets. According to the neighbored cells, the gene expressions of spatial transcriptomics cells are imputed through a graph neural network, where nodes are genes, and edges are based on cosine similarity between the ESM-2 embeddings of the gene-encoding proteins. The gene prediction uncertainty is further measured through a deep learning model. stImpute was shown to consistently outperform state-of-the-art methods across multiple datasets concerning imputation and clustering. stImpute also demonstrates robustness in producing consistent results that are insensitive to model parameters.

ScEMB: Learning context representation of genes based on large-scale single-cell transcriptomics.

The rapid advancement of single-cell transcriptomic technologies has led to the curation of millions of cellular profiles, providing unprecedented insights into cellular heterogeneity across various tissues and developmental stages. This growing wealth of data presents an opportunity to uncover complex gene-gene relationships, yet also poses significant computational challenges. We present scEMB, a transformer-based deep learning model developed to capture context-aware gene embeddings from large-scale single-cell transcriptomics data. Trained on over 30 million single-cell transcriptomes, scEMB utilizes an innovative binning strategy that integrates data across multiple platforms, effectively preserving both gene expression hierarchies and cell-type specificity. In downstream tasks such as batch integration, clustering, and cell type annotation, scEMB demonstrates superior performance compared to existing models like scGPT and Geneformer. Notably, scEMB excels in silico correlation analysis, accurately predicting gene perturbation effects in CRISPR-edited datasets and microglia state transition, identifying a few known Alzheimer's disease (AD) risks genes in top gene list. Additionally, scEMB offers robust fine-tuning capabilities for domain-specific applications, making it a versatile tool for tackling diverse biological problems such as therapeutic target discovery and disease modeling. scEMB represents a powerful tool for extracting biologically meaningful insights from complex gene expression data. Its ability to model in silico perturbation effects and conduct correlation analyses in the embedding space highlights its potential to accelerate discoveries in precision medicine and therapeutic development.

Loss of coordination between basic cellular processes in human aging.

Age-related loss of gene expression coordination has been reported for distinct cell types and may lead to impaired cellular function. Here we propose a method for quantifying age-related changes in transcriptional regulatory relationships between genes, based on a model learned from external data. We used this method to uncover age-related trends in gene-gene relationships across eight human tissues, which demonstrates that reduced co-expression may also result from coordinated transcriptional responses. Our analyses reveal similar numbers of strengthening and weakening gene-gene relationships with age, impacting both tissue-specific (for example, coagulation in blood) and ubiquitous biological functions. Regulatory relationships becoming weaker with age were established mostly between genes operating in distinct cellular processes. As opposed to that, regulatory relationships becoming stronger with age were established both within and between different cellular functions. Our work reveals that, although most transcriptional regulatory gene-gene relationships are maintained during aging, those with declining regulatory coupling result mostly from a loss of coordination between distinct cellular processes.

Gene regulatory networks in disease and ageing.

The precise control of gene expression is required for the maintenance of cellular homeostasis and proper cellular function, and the declining control of gene expression with age is considered a major contributor to age-associated changes in cellular physiology and disease. The coordination of gene expression can be represented through models of the molecular interactions that govern gene expression levels, so-called gene regulatory networks. Gene regulatory networks can represent interactions that occur through signal transduction, those that involve regulatory transcription factors, or statistical models of gene-gene relationships based on the premise that certain sets of genes tend to be coexpressed across a range of conditions and cell types. Advances in experimental and computational technologies have enabled the inference of these networks on an unprecedented scale and at unprecedented precision. Here, we delineate different types of gene regulatory networks and their cell-biological interpretation. We describe methods for inferring such networks from large-scale, multi-omics datasets and present applications that have aided our understanding of cellular ageing and disease mechanisms.

Gene-gene functional relationships in Alzheimer's disease: CELF1 regulates KLC1 alternative splicing

The causes of Alzheimer's disease (AD) are poorly understood, although many genes are known to be involved in this pathology. To gain insights into the underlying molecular mechanisms, it is essential to identify the relationships between individual AD genes. Previous work has shown that the splice variant E of KLC1 (KLC1_vE) promotes AD, and that the CELF1 gene, which encodes an RNA-binding protein involved in splicing regulation, is at a risk locus for AD. Here, we identified a functional link between CELF1 and KLC1 in AD pathogenesis. Transcriptomic data from human samples from different ethnic groups revealed that CELF1 mRNA levels are low in AD brains, and the splicing pattern of KLC1 is strongly correlated with CELF1 expression levels. Specifically, KLC1_vE is negatively correlated with CELF1. Depletion and overexpression experiments in cultured cells demonstrated that the CELF1 protein down-regulates KLC1_vE. In a cross-linking and immunoprecipitation sequencing (CLIP-seq) database, CELF1 directly binds to KLC1 RNA, following which it likely modulates terminal exon usage, hence KLC1_vE formation. These findings reveal a new pathogenic pathway where a risk allele of CELF1 is associated with reduced CELF1 expression, which up-regulates KLC1_vE to promote AD.

A muti-modal feature fusion method based on deep learning for predicting immunotherapy response

Immune checkpoint therapy (ICT) has greatly improved the survival of cancer patients in the past few years, but only a small number of patients respond to ICT. To predict ICT response, we developed a multi-modal feature fusion model based on deep learning (MFMDL). This model utilizes graph neural networks to map gene-gene relationships in gene networks to low dimensional vector spaces, and then fuses biological pathway features and immune cell infiltration features to make robust predictions of ICT. We used five datasets to validate the predictive performance of the MFMDL. These five datasets span multiple types of cancer, including melanoma, lung cancer, and gastric cancer. We found that the prediction performance of multi-modal feature fusion model based on deep learning is superior to other traditional ICT biomarkers, such as ICT targets or tumor microenvironment-associated markers. In addition, we also conducted ablation experiments to demonstrate the necessity of fusing different modal features, which can improve the prediction accuracy of the model.

Single-cell RNA sequencing data imputation using bi-level feature propagation.

Single-cell RNA sequencing (scRNA-seq) enables the exploration of cellular heterogeneity by analyzing gene expression profiles in complex tissues. However, scRNA-seq data often suffer from technical noise, dropout events and sparsity, hindering downstream analyses. Although existing works attempt to mitigate these issues by utilizing graph structures for data denoising, they involve the risk of propagating noise and fall short of fully leveraging the inherent data relationships, relying mainly on one of cell-cell or gene-gene associations and graphs constructed by initial noisy data. To this end, this study presents single-cell bilevel feature propagation (scBFP), two-step graph-based feature propagation method. It initially imputes zero values using non-zero values, ensuring that the imputation process does not affect the non-zero values due to dropout. Subsequently, it denoises the entire dataset by leveraging gene-gene and cell-cell relationships in the respective steps. Extensive experimental results on scRNA-seq data demonstrate the effectiveness of scBFP in various downstream tasks, uncovering valuable biological insights.

Abstract 7133: Phenomics-enabled discovery and optimization of small-molecule RBM39 degraders as an alternative to CDK12 targeting in high-grade serous ovarian cancer (HGSOC)

Abstract Phenomics-enabled drug discovery is a powerful approach to identify novel targets and relationships previously unappreciated in biology. We built a phenomics platform leveraging high content microscopy and generated unique phenoprints for >7,000 genes using CRISPR/Cas9 technology and a diverse chemical library of >1 million compounds. This phenomics dataset contains an unprecedented quantity of gene-gene, gene-compound, and compound-compound relationships. We applied our phenomics platform and an inference based target-agnostic drug discovery approach to discover novel genes and small molecules that mimic CDK12 inhibition, an important transcriptional regulator of DNA damage response, while avoiding inhibition of the paralog CDK13. We discovered a novel and unappreciated association between CDK12 and RBM39 as well as small molecule degraders of RBM39. We established structure activity relationships (SAR) solely leveraging our phenomics platform, to generate REC-1170204, an RBM39 degrader. Unlike inhibitors of CDK12, we show that REC-1170204 does not directly inhibit CDK12, CDK13 or other kinases. However, REC-1170204 demonstrates candidate quality properties with improved potency, selectivity, and drug-like characteristics. Finally we show that REC-1170204 shows efficacy in High Grade Serous Ovarian cancer (HGSOC) pre-clinical models and synergizes with PARP inhibition in a PARP-resistant patient-derived xenograft (PDX) model. Together, our data suggests that targeting RBM39 may provide an attractive and safer approach to targeting CDK12 in the clinic. Citation Format: Chase Neumann, Harish Shankaran, Kiran Nadella, Kelly Biette, Shane Rowley, Ethan Gardner, Shadi Swaidani, Vamshi Manda, Lu Chen, Daria Beshnova, Ashraf Saeed, Christopher Bailey, Janet Paulsen, Paul Rearden, Carl Brooks, Ashish Bhandari, Chris Gibson, Laura Schaevitz, Imran Haque, Hayley Donnella, Michael Cuccarese, Marie Evangelista. Phenomics-enabled discovery and optimization of small-molecule RBM39 degraders as an alternative to CDK12 targeting in high-grade serous ovarian cancer (HGSOC) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7133.

Abstract B049: Pancreatic beta cell stress pathways drive pancreatic ductal adenocarcinoma development in obesity

Abstract Epidemiologic studies have shown that obesity is a risk factor for more than a dozen cancers, including pancreatic ductal adenocarcinoma (PDAC). As PDAC is one of the leading causes of cancer-related death and obesity rates are rapidly increasing worldwide, it is imperative that we understand the mechanisms by which obesity promotes PDAC progression. To address this gap in knowledge, our lab combined a well-established genetic model for obesity (O: Lepob/ob) with a Kras-driven pancreatic cancer model (KC: Pdx1-Cre; KrasLSL-G12D/+) that mimics human PDAC progression. Resultant KCO mice had a drastic increase in tumor development compared to their lean counterparts, whereas induced weight loss intercepted tumor progression, showing that the tumor-promoting effects of obesity can be reversed. Bulk RNA sequencing of tumors from these mice revealed aberrantly increased expression of multiple peptide hormones expressed in insulin-producing beta (b) cells of pancreatic islets, including cholecystokinin (CCK) and amylin (Iapp). Strikingly, transgenic expression of CCK and Iapp in b-cells was sufficient to promote Kras-driven pancreatic tumorigenesis in non-obese mice. Single-cell RNA sequencing of b cells showed a trajectory of changes in gene expression with increasing obesity from wildtype to high-fat diet-fed to Lepob/ob mice. Further in silico analysis combining archetypal (AANet) with pseudotime (TrajectoryNet) methods demonstrated that b cells misexpressing hormones (CCK and Iapp) arose from a specific subtype of preexisting b cells, specifically virgin b cells, a rare, immature, regenerative population. Preliminary in vivo lineage tracing analyses in MIP-CreERT;LSL-TdTomato;Lepob/ob mice support these findings. Finally, we investigated the transcriptional networks that lead to aberrant CCK and Iapp expression in b cells. By combining Grainger causality analysis with known gene-gene relationships in the Transcriptional Regulatory Relationship Unraveled by Sentence-based Text mining (TRRUST) v2 database, we discovered a stress-induced mitogen-activated protein kinase (MAPK8)-cJun pathway that promotes pro-tumorigenic hormone expression in b cells. These combined experimental and computational methods nominate novel b cell targets to subvert the development of PDAC, an exocrine tumor. Citation Format: Cathy C. Garcia, Aarthi Venkat, Alex Tong, Sherry Agabiti, Lauren Lawres, Rebecca Cardone, Richard Kibbey, Smita Krishnaswamy, Mandar D Muzumdar. Pancreatic beta cell stress pathways drive pancreatic ductal adenocarcinoma development in obesity [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Pancreatic Cancer; 2023 Sep 27-30; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(2 Suppl):Abstract nr B049.

Exploring large-scale gene coexpression networks in peach (Prunus persica L.): a new tool for predicting gene function.

Peach is a model for Prunus genetics and genomics, however, identifying and validating genes associated to peach breeding traits is a complex task. A gene coexpression network (GCN) capable of capturing stable gene-gene relationships would help researchers overcome the intrinsic limitations of peach genetics and genomics approaches and outline future research opportunities. In this study, we created four GCNs from 604 Illumina RNA-Seq libraries. We evaluated the performance of every GCN in predicting functional annotations using an algorithm based on the 'guilty-by-association' principle. The GCN with the best performance was COO300, encompassing 21 956 genes. To validate its performance predicting gene function, we performed two case studies. In case study 1, we used two genes involved in fruit flesh softening: the endopolygalacturonases PpPG21 and PpPG22. Genes coexpressing with both genes were extracted and referred to as melting flesh (MF) network. Finally, we performed an enrichment analysis of MF network and compared the results with the current knowledge regarding peach fruit softening. The MF network mostly included genes involved in cell wall expansion and remodeling, and with expressions triggered by ripening-related phytohormones, such as ethylene, auxin, and methyl jasmonate. In case study 2, we explored potential targets of the anthocyanin regulator PpMYB10.1 by comparing its gene-centered coexpression network with that of its grapevine orthologues, identifying a common regulatory network. These results validated COO300 as a powerful tool for peach and Prunus research. This network, renamed as PeachGCN v1.0, and the scripts required to perform a function prediction analysis are available at https://github.com/felipecobos/PeachGCN.

Detecting Boolean Asymmetric Relationships with a Loop Counting Technique and its Implications for Analyzing Heterogeneity within Gene Expression Datasets.

Many traditional methods for analyzing gene-gene relationships focus on positive and negative correlations, both of which are a kind of 'symmetric' relationship. Biclustering is one such technique that typically searches for subsets of genes exhibiting correlated expression among a subset of samples. However, genes can also exhibit 'asymmetric' relationships, such as 'if-then' relationships used in boolean circuits. In this paper we develop a very general method that can be used to detect biclusters within gene-expression data that involve subsets of genes which are enriched for these 'boolean-asymmetric' relationships (BARs). These BAR-biclusters can correspond to heterogeneity that is driven by asymmetric gene-gene interactions, e.g., reflecting regulatory effects of one gene on another, rather than more standard symmetric interactions. Unlike typical approaches that search for BARs across the entire population, BAR-biclusters can detect asymmetric interactions that only occur among a subset of samples. We apply our method to a single-cell RNA-sequencing data-set, demonstrating that the statistically-significant BARbiclusters indeed contain additional information not present within the more traditional 'boolean-symmetric'-biclusters. For example, the BAR-biclusters involve different subsets of cells, and highlight different gene-pathways within the data-set. Moreover, by combining the boolean-asymmetric- and boolean-symmetricsignals, one can build linear classifiers which outperform those built using only traditional boolean-symmetric signals.

Contingency, repeatability, and predictability in the evolution of a prokaryotic pangenome

Pangenomes exhibit remarkable variability in many prokaryotic species, much of which is maintained through the processes of horizontal gene transfer and gene loss. Repeated acquisitions of near-identical homologs can easily be observed across pangenomes, leading to the question of whether these parallel events potentiate similar evolutionary trajectories, or whether the remarkably different genetic backgrounds of the recipients mean that postacquisition evolutionary trajectories end up being quite different. In this study, we present a machine learning method that predicts the presence or absence of genes in the Escherichia coli pangenome based on complex patterns of the presence or absence of other accessory genes within a genome. Our analysis leverages the repeated transfer of genes through the E. coli pangenome to observe patterns of repeated evolution following similar events. We find that the presence or absence of a substantial set of genes is highly predictable from other genes alone, indicating that selection potentiates and maintains gene-gene co-occurrence and avoidance relationships deterministically over long-term bacterial evolution and is robust to differences in host evolutionary history. We propose that at least part of the pangenome can be understood as a set of genes with relationships that govern their likely cohabitants, analogous to an ecosystem's set of interacting organisms. Our findings indicate that intragenomic gene fitness effects may be key drivers of prokaryotic evolution, influencing the repeated emergence of complex gene-gene relationships across the pangenome.

Literature-based predictions of Mendelian disease therapies

In the effort to treat Mendelian disorders, correcting the underlying molecular imbalance may be more effective than symptomatic treatment. Identifying treatments that might accomplish this goal requires extensive and up-to-date knowledge of molecular pathways-including drug-gene and gene-gene relationships. To address this challenge, we present "parsing modifiers via article annotations" (PARMESAN), a computational tool that searches PubMed and PubMed Central for information to assemble these relationships into a central knowledge base. PARMESAN then predicts putatively novel drug-gene relationships, assigning an evidence-based score to each prediction. We compare PARMESAN's drug-gene predictions to all of the drug-gene relationships displayed by the Drug-Gene Interaction Database (DGIdb) and show that higher-scoring relationship predictions are more likely to match the directionality (up- versus down-regulation) indicated by this database. PARMESAN had more than 200,000 drug predictions scoring above 8 (as one example cutoff), for more than 3,700 genes. Among these predicted relationships, 210 were registered in DGIdb and 201 (96%) had matching directionality. This publicly available tool provides an automated way to prioritize drug screens to target the most-promising drugs to test, thereby saving time and resources in the development of therapeutics for genetic disorders.

Homogeneous Space Construction and Projection for Single-Cell Expression Prediction Based on Deep Learning.

Predicting cellular responses to perturbations is an unsolved problem in biology. Traditional approaches assume that different cell types respond similarly to perturbations. However, this assumption does not take into account the context of genome interactions in different cell types, which leads to compromised prediction quality. More recently, deep learning models used to discover gene-gene relationships can yield more accurate predictions of cellular responses. The huge difference in biological information between different cell types makes it difficult for deep learning models to encode data into a continuous low-dimensional feature space, which means that the features captured by the latent space may not be continuous. Therefore, the mapping relationship between the two conditional spaces learned by the model can only be applied where the real reference data resides, leading to the wrong mapping of the predicted target cells because they are not in the same domain as the reference data. In this paper, we propose an information-navigated variational autoencoder (INVAE), a deep neural network for cell perturbation response prediction. INVAE filters out information that is not conducive to predictive performance. For the remaining information, INVAE constructs a homogeneous space of control conditions, and finds the mapping relationship between the control condition space and the perturbation condition space. By embedding the target unit into the control space and then mapping it to the perturbation space, we can predict the perturbed state of the target unit. Comparing our proposed method with other three state-of-the-art methods on three real datasets, experimental results show that INVAE outperforms existing methods in cell state prediction after perturbation. Furthermore, we demonstrate that filtering out useless information not only improves prediction accuracy but also reveals similarities in how genes in different cell types are regulated following perturbation.

Context-defined cancer co-dependency mapping identifies a functional interplay between PRC2 and MLL-MEN1 complex in lymphoma

Interplay between chromatin-associated complexes and modifications critically contribute to the partitioning of epigenome into stable and functionally distinct domains. Yet there is a lack of systematic identification of chromatin crosstalk mechanisms, limiting our understanding of the dynamic transition between chromatin states during development and disease. Here we perform co-dependency mapping of genes using CRISPR-Cas9-mediated fitness screens in pan-cancer cell lines to quantify gene-gene functional relationships. We identify 145 co-dependency modules and further define the molecular context underlying the essentiality of these modules by incorporating mutational, epigenome, gene expression and drug sensitivity profiles of cell lines. These analyses assign new protein complex composition and function, and predict new functional interactions, including an unexpected co-dependency between two transcriptionally counteracting chromatin complexes - polycomb repressive complex 2 (PRC2) and MLL-MEN1 complex. We show that PRC2-mediated H3K27 tri-methylation regulates the genome-wide distribution of MLL1 and MEN1. In lymphoma cells with EZH2 gain-of-function mutations, the re-localization of MLL-MEN1 complex drives oncogenic gene expression and results in a hypersensitivity to pharmacologic inhibition of MEN1. Together, our findings provide a resource for discovery of trans-regulatory interactions as mechanisms of chromatin regulation and potential targets of synthetic lethality.

ACDC: a general approach for detecting phenotype or exposure associated co-expression.

Existing module-based differential co-expression methods identify differences in gene-gene relationships across phenotype or exposure structures by testing for consistent changes in transcription abundance. Current methods only allow for assessment of co-expression variation across a singular, binary or categorical exposure or phenotype, limiting the information that can be obtained from these analyses. Here, we propose a novel approach for detection of differential co-expression that simultaneously accommodates multiple phenotypes or exposures with binary, ordinal, or continuous data types. We report an application to two cohorts of asthmatic patients with varying levels of asthma control to identify associations between gene co-expression and asthma control test scores. Results suggest that both expression levels and covariances of ADORA3, ALOX15, and IDO1 are associated with asthma control. ACDC is a flexible extension to existing methodology that can detect differential co-expression across varying external variables.

Mapping the gene space at single-cell resolution with gene signal pattern analysis

Abstract In single-cell RNA sequencing analysis, several computational methods have been developed to map the cellular state space, but little has been done to map the gene space. A mapping that preserves gene-gene relationships within the dataset is particularly useful for characterizing cellular heterogeneity within cell types, where boundaries between cell subpopulations are often unclear or even arbitrary. Here, we present gene signal pattern analysis, a new paradigm for analyzing single cells. We build a cell-cell graph and design a dictionary of diffusion wavelets, capturing a multiscale view of the cell space. We then transform genes by the dictionary and learn a reduced gene representation. Given the gap in prior research for this problem, we design nine alternative strategies and three benchmarks for evaluating preservation of gene-gene relationships, all of which are outperformed by diffusion wavelet-transformed signals. We also define, calculate, and evaluate localization, a key property of a gene signal on the cellular graph. We demonstrate the utility of gene signal pattern analysis on T cells from a mouse model of peripheral tolerance in skin. The gene space mapping reveals a continuum of gene signals characterized by T cell subtypes and transcriptional programs related to effector function and proliferation. Furthermore, we built a multiscale manifold of 48 melanoma patient samples, demonstrating the ability of our method to characterize differences between responders and non-responders to checkpoint immunotherapy. Together, we show gene signal pattern analysis, through methodology from graph signal processing, spectral graph theory, and machine learning, represents an avenue for future research in scRNA-seq analysis. Gruber Foundation

157 Identifying Spatial Transcriptomics Signaling Networks in Human Glioblastoma Using Graph-Based Machine Learning

INTRODUCTION: The median survival for glioblastoma multiforme (GBM) has only improved by 4.6 months in more than 100 years, underlying the need for innovative methods for understanding this devastating tumor. Spatial transcriptomics is a new method of whole-transcriptome assessment which allows for simultaneous visualization of mRNA expression in situ with histopathology. METHODS: We applied spARC to 25,517 Unique Molecular Identifiers from 10 surgically resected human IDH-mutant astrocytic gliomas (Grade II (n = 3) and Grade III (n = 3)) and GBMs (n = 4). spARC uses manifold learning, graph filters, data diffusion and diffusion filtration to enable discovery of gene-gene relationships, spatial co-localization of genes, and receptor-ligand interactions applicable across a variety of spatial transcriptomic methods. A Neuropathologist annotated whole-slide images blind to our computational analysis and we used in-situ hybridization and immunohistochemistry to demonstrate that spaRC imputation recovery of gene expression is superior to raw expression. RESULTS: We demonstrate that spARC can identify spatially restricted tumor niches of dense tumor, leading edge, infiltration/high tumor burden, hemoglobin, and vasculature with high specificity. These clusters can be identified with genes including CD44, VIM, LDHA (dense tumor); SNAP25, UCHL1, OLMF1 (leading edge); OLIG2, MBP, PDGFRA (Infiltrating/High Tumor Burden). In our assessment of ligand-receptor interactions we found that GBM samples were enriched for SPP1-CD44 signaling. We validated this using GBM TCGA data stratifying patients based on expression of SPP1 and CD44 genes and found that high expression conferred lower survival at 12 months. CONCLUSIONS: Spatial transcriptomics paired with graph-based machine learning methods applied to human GBMs is a powerful technique able to uncover signaling networks relevant to patient survival and may aid in developing new therapeutic strategies.

Elucidating dynamic cell lineages and gene networks in time-course single cell differentiation

Single cell RNA sequencing (scRNA-seq) technologies provide researchers with an unprecedented opportunity to exploit cell heterogeneity. For example, the sequenced cells belong to various cell lineages, which may have different cell fates in stem and progenitor cells. Those cells may differentiate into various mature cell types in a cell differentiation process. To trace the behavior of cell differentiation, researchers reconstruct cell lineages and predict cell fates by ordering cells chronologically into a trajectory with a pseudo-time. However, in scRNA-seq experiments, there are no cell-to-cell correspondences along with the time to reconstruct the cell lineages, which creates a significant challenge for cell lineage tracing and cell fate prediction. Therefore, methods that can accurately reconstruct the dynamic cell lineages and predict cell fates are highly desirable.In this article, we develop an innovative machine-learning framework called Cell Smoothing Transformation (CellST) to elucidate the dynamic cell fate paths and construct gene networks in cell differentiation processes. Unlike the existing methods that construct one single bulk cell trajectory, CellST builds cell trajectories and tracks behaviors for each individual cell. Additionally, CellST can predict cell fates even for less frequent cell types. Based on the individual cell fate trajectories, CellST can further construct dynamic gene networks to model gene-gene relationships along the cell differentiation process and discover critical genes that potentially regulate cells into various mature cell types.

Abstract A015: Cancer co-dependency mapping identifies a functional interplay between PRC2 and MLL-MEN1 complex in lymphoma

Abstract Interplays between chromatin-associated complexes and modifications critically contribute to the partitioning of epigenome into stable and functionally distinct domains. Yet there is a lack of systematic identification of chromatin crosstalk mechanisms, limiting our understanding of the dynamic transition between chromatin states during development and disease. Here we perform co-dependency mapping of genes encoding nuclear proteins using CRISPR-Cas9-mediated fitness screens in pan-cancer cell lines to quantify gene-gene functional relationships. We identify 84 co-dependency modules and further define the molecular context underlying the essentiality of these modules by incorporating mutational, epigenome, gene expression and drug sensitivity profiles of cell lines. These analyses assign new protein complex functions and biochemical interactions, and predict new chromatin crosstalk, including an unexpected co-dependency between two transcriptionally counteracting chromatin complexes - polycomb repressive complex 2 (PRC2) and MLL-MEN1 complex. We show that PRC2-mediated H3K27 tri-methylation regulates the genome-wide distribution of MLL1 and MEN1. In lymphoma cells with EZH2 gain-of-function mutations, the re-localization of MLL-MEN1 complex drives oncogenic gene expression and results in a hypersensitivity to pharmacologic inhibition of MEN1. Together, our findings provide a resource for discovery of trans-regulatory interactions as mechanisms of chromatin regulation and potential targets of synthetic lethality. Citation Format: Xiao Chen, Yinglu Li, Chao Lu. Cancer co-dependency mapping identifies a functional interplay between PRC2 and MLL-MEN1 complex in lymphoma. [abstract]. In: Proceedings of the AACR Special Conference: Cancer Epigenomics; 2022 Oct 6-8; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2022;82(23 Suppl_2):Abstract nr A015.