Single-cell gene regulatory network prediction by explainable AI

Decoding pan-cancer treatment outcomes using multimodal real-world data and explainable artificial intelligence

Despite advances in precision oncology, clinical decision-making still relies on limited variables and expert knowledge. To address this limitation, we combined multimodal real-world data and explainable artificial intelligence (xAI) to introduce AI-derived (AID) markers for clinical decision support. We used xAI to decode the outcome of 15,726 patients across 38 solid cancer entities based on 350 markers, including clinical records, image-derived body compositions, and mutational tumor profiles. xAI determined the prognostic contribution of each clinical marker at the patient level and identified 114 key markers that accounted for 90% of the neural network's decision process. Moreover, xAI enabled us to uncover 1,373 prognostic interactions between markers. Our approach was validated in an independent cohort of 3,288 patients with lung cancer from a US nationwide electronic health record-derived database. These results show the potential of xAI to transform the assessment of clinical variables and enable personalized, data-driven cancer care.

BACH1 as a key driver in rheumatoid arthritis fibroblast-like synoviocytes identified through gene network analysis

RNA-sequencing and differential gene expression studies have significantly advanced our understanding of pathogenic pathways underlying rheumatoid arthritis (RA). Yet, little is known about cell-specific regulatory networks and their contributions to disease. In this study, we focused on fibroblast-like synoviocytes (FLS), a cell type central to disease pathogenesis and joint damage in RA. We used a strategy that computed sample-specific gene regulatory networks to compare network properties between RA and osteoarthritis FLS. We identified 28 transcription factors (TFs) as key regulators central to the signatures of RA FLS. Six of these TFs are new and have not been previously implicated in RA through ex vivo or in vivo studies, and included BACH1, HLX, and TGIF1. Several of these TFs were found to be co-regulated, and BACH1 emerged as the most significant TF and regulator. The main BACH1 targets included those implicated in fatty acid metabolism and ferroptosis. The discovery of BACH1 was validated in experiments with RA FLS. Knockdown of BACH1 in RA FLS significantly affected the gene expression signatures, reduced cell adhesion and mobility, interfered with the formation of thick actin fibers, and prevented the polarized formation of lamellipodia, all required for the RA destructive behavior of FLS. This study establishes BACH1 as a central regulator of RA FLS phenotypes and suggests its potential as a therapeutic target to selectively modulate RA FLS.

OneSC: a computational platform for recapitulating cell state transitions

MOTIVATION

Computational modeling of cell state transitions has been a great interest of many in the field of developmental biology, cancer biology, and cell fate engineering because it enables performing perturbation experiments in silico more rapidly and cheaply than could be achieved in a lab. Recent advancements in single-cell RNA-sequencing (scRNA-seq) allow the capture of high-resolution snapshots of cell states as they transition along temporal trajectories. Using these high-throughput datasets, we can train computational models to generate in silico "synthetic" cells that faithfully mimic the temporal trajectories.

RESULTS

Here we present OneSC, a platform that can simulate cell state transitions using systems of stochastic differential equations govern by a regulatory network of core transcription factors (TFs). Different from many current network inference methods, OneSC prioritizes on generating Boolean network that produces faithful cell state transitions and terminal cell states that mimic real biological systems. Applying OneSC to real data, we inferred a core TF network using a mouse myeloid progenitor scRNA-seq dataset and showed that the dynamical simulations of that network generate synthetic single-cell expression profiles that faithfully recapitulate the four myeloid differentiation trajectories going into differentiated cell states (erythrocytes, megakaryocytes, granulocytes, and monocytes). Finally, through the in silico perturbations of the mouse myeloid progenitor core network, we showed that OneSC can accurately predict cell fate decision biases of TF perturbations that closely match with previous experimental observations.

AVAILABILITY AND IMPLEMENTATION

OneSC is implemented as a Python package on GitHub (https://github.com/CahanLab/oneSC) and on Zenodo (https://zenodo.org/records/14052421).

AutoXAI4Omics: an automated explainable AI tool for omics and tabular data

Machine learning (ML) methods offer opportunities for gaining insights into the intricate workings of complex biological systems, and their applications are increasingly prominent in the analysis of omics data to facilitate tasks, such as the identification of novel biomarkers and predictive modeling of phenotypes. For scientists and domain experts, leveraging user-friendly ML pipelines can be incredibly valuable, enabling them to run sophisticated, robust, and interpretable models without requiring in-depth expertise in coding or algorithmic optimization. By streamlining the process of model development and training, researchers can devote their time and energies to the critical tasks of biological interpretation and validation, thereby maximizing the scientific impact of ML-driven insights. Here, we present an entirely automated open-source explainable AI tool, AutoXAI4Omics, that performs classification and regression tasks from omics and tabular numerical data. AutoXAI4Omics accelerates scientific discovery by automating processes and decisions made by AI experts, e.g. selection of the best feature set, hyper-tuning of different ML algorithms and selection of the best ML model for a specific task and dataset. Prior to ML analysis AutoXAI4Omics incorporates feature filtering options that are tailored to specific omic data types. Moreover, the insights into the predictions that are provided by the tool through explainability analysis highlight associations between omic feature values and the targets under investigation, e.g. predicted phenotypes, facilitating the identification of novel actionable insights. AutoXAI4Omics is available at: https://github.com/IBM/AutoXAI4Omics.

Cell-specific gene networks and drivers in rheumatoid arthritis synovial tissues

Rheumatoid arthritis (RA) is a common autoimmune and inflammatory disease characterized by inflammation and hyperplasia of the synovial tissues. RA pathogenesis involves multiple cell types, genes, transcription factors (TFs) and networks. Yet, little is known about the TFs, and key drivers and networks regulating cell function and disease at the synovial tissue level, which is the site of disease. In the present study, we used available RNA-seq databases generated from synovial tissues and developed a novel approach to elucidate cell type-specific regulatory networks on synovial tissue genes in RA. We leverage established computational methodologies to infer sample-specific gene regulatory networks and applied statistical methods to compare network properties across phenotypic groups (RA versus osteoarthritis). We developed computational approaches to rank TFs based on their contribution to the observed phenotypic differences between RA and controls across different cell types. We identified 18 (fibroblast-like synoviocyte), 16 (T cells), 19 (B cells) and 11 (monocyte) key regulators in RA synovial tissues. Interestingly, fibroblast-like synoviocyte (FLS) and B cells were driven by multiple independent co-regulatory TF clusters that included MITF, HLX, BACH1 (FLS) and KLF13, FOSB, FOSL1 (B cells). However, monocytes were collectively governed by a single cluster of TF drivers, responsible for the main phenotypic differences between RA and controls, which included RFX5, IRF9, CREB5. Among several cell subset and pathway changes, we also detected reduced presence of Natural killer T (NKT) cells and eosinophils in RA synovial tissues. Overall, our novel approach identified new and previously unsuspected Key driver genes (KDG), TF and networks and should help better understanding individual cell regulation and co-regulatory networks in RA pathogenesis, as well as potentially generate new targets for treatment.

Designing interpretable deep learning applications for functional genomics: a quantitative analysis

Deep learning applications have had a profound impact on many scientific fields, including functional genomics. Deep learning models can learn complex interactions between and within omics data; however, interpreting and explaining these models can be challenging. Interpretability is essential not only to help progress our understanding of the biological mechanisms underlying traits and diseases but also for establishing trust in these model's efficacy for healthcare applications. Recognizing this importance, recent years have seen the development of numerous diverse interpretability strategies, making it increasingly difficult to navigate the field. In this review, we present a quantitative analysis of the challenges arising when designing interpretable deep learning solutions in functional genomics. We explore design choices related to the characteristics of genomics data, the neural network architectures applied, and strategies for interpretation. By quantifying the current state of the field with a predefined set of criteria, we find the most frequent solutions, highlight exceptional examples, and identify unexplored opportunities for developing interpretable deep learning models in genomics.

DeepIMAGER: Deeply Analyzing Gene Regulatory Networks from scRNA-seq Data

Understanding the dynamics of gene regulatory networks (GRNs) across diverse cell types poses a challenge yet holds immense value in unraveling the molecular mechanisms governing cellular processes. Current computational methods, which rely solely on expression changes from bulk RNA-seq and/or scRNA-seq data, often result in high rates of false positives and low precision. Here, we introduce an advanced computational tool, DeepIMAGER, for inferring cell-specific GRNs through deep learning and data integration. DeepIMAGER employs a supervised approach that transforms the co-expression patterns of gene pairs into image-like representations and leverages transcription factor (TF) binding information for model training. It is trained using comprehensive datasets that encompass scRNA-seq profiles and ChIP-seq data, capturing TF-gene pair information across various cell types. Comprehensive validations on six cell lines show DeepIMAGER exhibits superior performance in ten popular GRN inference tools and has remarkable robustness against dropout-zero events. DeepIMAGER was applied to scRNA-seq datasets of multiple myeloma (MM) and detected potential GRNs for TFs of RORC, MITF, and FOXD2 in MM dendritic cells. This technical innovation, combined with its capability to accurately decode GRNs from scRNA-seq, establishes DeepIMAGER as a valuable tool for unraveling complex regulatory networks in various cell types.

Interpretable deep learning in single-cell omics

MOTIVATION

Single-cell omics technologies have enabled the quantification of molecular profiles in individual cells at an unparalleled resolution. Deep learning, a rapidly evolving sub-field of machine learning, has instilled a significant interest in single-cell omics research due to its remarkable success in analysing heterogeneous high-dimensional single-cell omics data. Nevertheless, the inherent multi-layer nonlinear architecture of deep learning models often makes them 'black boxes' as the reasoning behind predictions is often unknown and not transparent to the user. This has stimulated an increasing body of research for addressing the lack of interpretability in deep learning models, especially in single-cell omics data analyses, where the identification and understanding of molecular regulators are crucial for interpreting model predictions and directing downstream experimental validations.

RESULTS

In this work, we introduce the basics of single-cell omics technologies and the concept of interpretable deep learning. This is followed by a review of the recent interpretable deep learning models applied to various single-cell omics research. Lastly, we highlight the current limitations and discuss potential future directions.

OneSC: A computational platform for recapitulating cell state transitions

Computational modelling of cell state transitions has been a great interest of many in the field of developmental biology, cancer biology and cell fate engineering because it enables performing perturbation experiments in silico more rapidly and cheaply than could be achieved in a wet lab. Recent advancements in single-cell RNA sequencing (scRNA-seq) allow the capture of high-resolution snapshots of cell states as they transition along temporal trajectories. Using these high-throughput datasets, we can train computational models to generate in silico 'synthetic' cells that faithfully mimic the temporal trajectories. Here we present OneSC, a platform that can simulate synthetic cells across developmental trajectories using systems of stochastic differential equations govern by a core transcription factors (TFs) regulatory network. Different from the current network inference methods, OneSC prioritizes on generating Boolean network that produces faithful cell state transitions and steady cell states that mimic real biological systems. Applying OneSC to real data, we inferred a core TF network using a mouse myeloid progenitor scRNA-seq dataset and showed that the dynamical simulations of that network generate synthetic single-cell expression profiles that faithfully recapitulate the four myeloid differentiation trajectories going into differentiated cell states (erythrocytes, megakaryocytes, granulocytes and monocytes). Finally, through the in-silico perturbations of the mouse myeloid progenitor core network, we showed that OneSC can accurately predict cell fate decision biases of TF perturbations that closely match with previous experimental observations.

Unique challenges and best practices for single cell transcriptomic analysis in toxicology

The application and analysis of single-cell transcriptomics in toxicology presents unique challenges. These include identifying cell sub-populations sensitive to perturbation; interpreting dynamic shifts in cell type proportions in response to chemical exposures; and performing differential expression analysis in dose-response studies spanning multiple treatment conditions. This review examines these challenges while presenting best practices for critical single cell analysis tasks. This covers areas such as cell type identification; analysis of differential cell type abundance; differential gene expression; and cellular trajectories. Towards enhancing the use of single-cell transcriptomics in toxicology, this review aims to address key challenges in this field and offer practical analytical solutions. Overall, applying appropriate bioinformatic techniques to single-cell transcriptomic data can yield valuable insights into cellular responses to toxic exposures.

[Explainable artificial intelligence in pathology]

With the advancements in precision medicine, the demands on pathological diagnostics have increased, requiring standardized, quantitative, and integrated assessments of histomorphological and molecular pathological data. Great hopes are placed in artificial intelligence (AI) methods, which have demonstrated the ability to analyze complex clinical, histological, and molecular data for disease classification, biomarker quantification, and prognosis estimation. This paper provides an overview of the latest developments in pathology AI, discusses the limitations, particularly concerning the black box character of AI, and describes solutions to make decision processes more transparent using methods of so-called explainable AI (XAI).

Toward Explainable Artificial Intelligence for Precision Pathology

The rapid development of precision medicine in recent years has started to challenge diagnostic pathology with respect to its ability to analyze histological images and increasingly large molecular profiling data in a quantitative, integrative, and standardized way. Artificial intelligence (AI) and, more precisely, deep learning technologies have recently demonstrated the potential to facilitate complex data analysis tasks, including clinical, histological, and molecular data for disease classification; tissue biomarker quantification; and clinical outcome prediction. This review provides a general introduction to AI and describes recent developments with a focus on applications in diagnostic pathology and beyond. We explain limitations including the black-box character of conventional AI and describe solutions to make machine learning decisions more transparent with so-called explainable AI. The purpose of the review is to foster a mutual understanding of both the biomedical and the AI side. To that end, in addition to providing an overview of the relevant foundations in pathology and machine learning, we present worked-through examples for a better practical understanding of what AI can achieve and how it should be done.

Cell-Specific Gene Networks and Drivers in Rheumatoid Arthritis Synovial Tissues

Rheumatoid arthritis (RA) is a common autoimmune and inflammatory disease characterized by inflammation and hyperplasia of the synovial tissues. RA pathogenesis involves multiple cell types, genes, transcription factors (TFs) and networks. Yet, little is known about the TFs, and key drivers and networks regulating cell function and disease at the synovial tissue level, which is the site of disease. In the present study, we used available RNA-seq databases generated from synovial tissues and developed a novel approach to elucidate cell type-specific regulatory networks on synovial tissue genes in RA. We leverage established computational methodologies to infer sample-specific gene regulatory networks and applied statistical methods to compare network properties across phenotypic groups (RA versus osteoarthritis). We developed computational approaches to rank TFs based on their contribution to the observed phenotypic differences between RA and controls across different cell types. We identified 18,16,19,11 key regulators of fibroblast-like synoviocyte (FLS), T cells, B cells, and monocyte signatures and networks, respectively, in RA synovial tissues. Interestingly, FLS and B cells were driven by multiple independent co-regulatory TF clusters that included MITF, HLX, BACH1 (FLS) and KLF13, FOSB, FOSL1 (synovial B cells). However, monocytes were collectively governed by a single cluster of TF drivers, responsible for the main phenotypic differences between RA and controls, which included RFX5, IRF9, CREB5. Among several cell subset and pathway changes, we also detected reduced presence of NKT cell and eosinophils in RA synovial tissues. Overall, our novel approach identified new and previously unsuspected KDG, TF and networks and should help better understanding individual cell regulation and co-regulatory networks in RA pathogenesis, as well as potentially generate new targets for treatment.

Gene Network Analyses Identify Co-regulated Transcription Factors and BACH1 as a Key Driver in Rheumatoid Arthritis Fibroblast-like Synoviocytes

RNA-sequencing and differential gene expression studies have significantly advanced our understanding of pathogenic pathways underlying Rheumatoid Arthritis (RA). Yet, little is known about cell-specific regulatory networks and their contributions to disease. In this study, we focused on fibroblast-like synoviocytes (FLS), a cell type central to disease pathogenesis and joint damage in RA. We used a strategy that computed sample-specific gene regulatory networks (GRNs) to compare network properties between RA and osteoarthritis FLS. We identified 28 transcription factors (TFs) as key regulators central to the signatures of RA FLS. Six of these TFs are new and have not been previously implicated in RA, and included BACH1, HLX, and TGIF1. Several of these TFs were found to be co-regulated, and BACH1 emerged as the most significant TF and regulator. The main BACH1 targets included those implicated in fatty acid metabolism and ferroptosis. The discovery of BACH1 was validated in experiments with RA FLS. Knockdown of BACH1 in RA FLS significantly affected the gene expression signatures, reduced cell adhesion and mobility, interfered with the formation of thick actin fibers, and prevented the polarized formation of lamellipodia, all required for the RA destructive behavior of FLS. This is the first time that BACH1 is shown to have a central role in the regulation of FLS phenotypes, and gene expression signatures, as well as in ferroptosis and fatty acid metabolism. These new discoveries have the potential to become new targets for treatments aimed at selectively targeting the RA FLS.

Novel research and future prospects of artificial intelligence in cancer diagnosis and treatment

Research into the potential benefits of artificial intelligence for comprehending the intricate biology of cancer has grown as a result of the widespread use of deep learning and machine learning in the healthcare sector and the availability of highly specialized cancer datasets. Here, we review new artificial intelligence approaches and how they are being used in oncology. We describe how artificial intelligence might be used in the detection, prognosis, and administration of cancer treatments and introduce the use of the latest large language models such as ChatGPT in oncology clinics. We highlight artificial intelligence applications for omics data types, and we offer perspectives on how the various data types might be combined to create decision-support tools. We also evaluate the present constraints and challenges to applying artificial intelligence in precision oncology. Finally, we discuss how current challenges may be surmounted to make artificial intelligence useful in clinical settings in the future.

Advances and challenges in thyroid cancer: The interplay of genetic modulators, targeted therapies, and AI-driven approaches

Thyroid cancer continues to exhibit a rising incidence globally, predominantly affecting women. Despite stable mortality rates, the unique characteristics of thyroid carcinoma warrant a distinct approach. Differentiated thyroid cancer, comprising most cases, is effectively managed through standard treatments such as thyroidectomy and radioiodine therapy. However, rarer variants, including anaplastic thyroid carcinoma, necessitate specialized interventions, often employing targeted therapies. Although these drugs focus on symptom management, they are not curative. This review delves into the fundamental modulators of thyroid cancers, encompassing genetic, epigenetic, and non-coding RNA factors while exploring their intricate interplay and influence. Epigenetic modifications directly affect the expression of causal genes, while long non-coding RNAs impact the function and expression of micro-RNAs, culminating in tumorigenesis. Additionally, this article provides a concise overview of the advantages and disadvantages associated with pharmacological and non-pharmacological therapeutic interventions in thyroid cancer. Furthermore, with technological advancements, integrating modern software and computing into healthcare and medical practices has become increasingly prevalent. Artificial intelligence and machine learning techniques hold the potential to predict treatment outcomes, analyze data, and develop personalized therapeutic approaches catering to patient specificity. In thyroid cancer, cutting-edge machine learning and deep learning technologies analyze factors such as ultrasonography results for tumor textures and biopsy samples from fine needle aspirations, paving the way for a more accurate and effective therapeutic landscape in the near future.

Precision cancer medicine: Concepts, current practice, and future developments

Precision cancer medicine is a multidisciplinary team effort that requires involvement and commitment of many stakeholders including the society at large. Building on the success of significant advances in precision therapy for oncological patients over the last two decades, future developments will be significantly shaped by improvements in scalable molecular diagnostics in which increasingly complex multilayered datasets require transformation into clinically useful information guiding patient management at fast turnaround times. Adaptive profiling strategies involving tissue- and liquid-based testing that account for the immense plasticity of cancer during the patient's journey and also include early detection approaches are already finding their way into clinical routine and will become paramount. A second major driver is the development of smart clinical trials and trial concepts which, complemented by real-world evidence, rapidly broaden the spectrum of therapeutic options. Tight coordination with regulatory agencies and health technology assessment bodies is crucial in this context. Multicentric networks operating nationally and internationally are key in implementing precision oncology in clinical practice and support developing and improving the ecosystem and framework needed to turn invocation into benefits for patients. The review provides an overview of the diagnostic tools, innovative clinical studies, and collaborative efforts needed to realize precision cancer medicine.

Predicting gene regulatory links from single-cell RNA-seq data using graph neural networks

Single-cell RNA-sequencing (scRNA-seq) has emerged as a powerful technique for studying gene expression patterns at the single-cell level. Inferring gene regulatory networks (GRNs) from scRNA-seq data provides insight into cellular phenotypes from the genomic level. However, the high sparsity, noise and dropout events inherent in scRNA-seq data present challenges for GRN inference. In recent years, the dramatic increase in data on experimentally validated transcription factors binding to DNA has made it possible to infer GRNs by supervised methods. In this study, we address the problem of GRN inference by framing it as a graph link prediction task. In this paper, we propose a novel framework called GNNLink, which leverages known GRNs to deduce the potential regulatory interdependencies between genes. First, we preprocess the raw scRNA-seq data. Then, we introduce a graph convolutional network-based interaction graph encoder to effectively refine gene features by capturing interdependencies between nodes in the network. Finally, the inference of GRN is obtained by performing matrix completion operation on node features. The features obtained from model training can be applied to downstream tasks such as measuring similarity and inferring causality between gene pairs. To evaluate the performance of GNNLink, we compare it with six existing GRN reconstruction methods using seven scRNA-seq datasets. These datasets encompass diverse ground truth networks, including functional interaction networks, Loss of Function/Gain of Function data, non-specific ChIP-seq data and cell-type-specific ChIP-seq data. Our experimental results demonstrate that GNNLink achieves comparable or superior performance across these datasets, showcasing its robustness and accuracy. Furthermore, we observe consistent performance across datasets of varying scales. For reproducibility, we provide the data and source code of GNNLink on our GitHub repository: https://github.com/sdesignates/GNNLink.

[Molecular tumor diagnostics as the driving force behind precision oncology]

Molecular pathological diagnostics plays a central role in personalized oncology and requires multidisciplinary teamwork. It is just as relevant for the individual patient who is being treated with an approved therapy method or an individual treatment attempt as it is for prospective clinical studies that require the identification of specific therapeutic target structures or complex biomarkers for study inclusion. It is also of crucial importance for the generation of real-world data, which is becoming increasingly important for drug development. Future developments will be significantly shaped by improvements in scalable molecular diagnostics, in which increasingly complex and multi-layered data sets must be quickly converted into clinically useful information. One focus will be on the development of adaptive diagnostic strategies in order to be able to depict the enormous plasticity of a cancer disease over time.

Neural cell diversity in the light of single-cell transcriptomics

The development of highly parallel and affordable high-throughput single-cell transcriptomics technologies has revolutionized our understanding of brain complexity. These methods have been used to build cellular maps of the brain, its different regions, and catalog the diversity of cells in each of them during development, aging and even in disease. Now we know that cellular diversity is way beyond what was previously thought. Single-cell transcriptomics analyses have revealed that cell types previously considered homogeneous based on imaging techniques differ depending on several factors including sex, age and location within the brain. The expression profiles of these cells have also been exploited to understand which are the regulatory programs behind cellular diversity and decipher the transcriptional pathways driving them. In this review, we summarize how single-cell transcriptomics have changed our view on the cellular diversity in the human brain, and how it could impact the way we study neurodegenerative diseases. Moreover, we describe the new computational approaches that can be used to study cellular differentiation and gain insight into the functions of individual cell populations under different conditions and their alterations in disease.