Recovering false negatives in CRISPR fitness screens with JLOE.

It is widely accepted that pooled library CRISPR knockout screens offer greater sensitivity and specificity than prior technologies in detecting genes whose disruption leads to fitness defects, a critical step in identifying candidate cancer targets. However, the assumption that CRISPR screens are saturating has been largely untested. Through integrated analysis of screen data in cancer cell lines generated by the Cancer Dependency Map, we show that a typical CRISPR screen has a ∼20% false negative rate, in addition to library-specific false negatives. Replicability falls sharply as gene expression decreases, while cancer subtype-specific genes within a tissue show distinct profiles compared to false negatives. Cumulative analyses across tissues improves our understanding of core essential genes and suggest only a small number of lineage-specific essential genes, enriched for transcription factors that define pathways of tissue differentiation. To recover false negatives, we introduce a method, Joint Log Odds of Essentiality (JLOE), which builds on our prior work with BAGEL to selectively rescue the false negatives without an increased false discovery rate.

Forecasting acute kidney injury and resource utilization in ICU patients using longitudinal, multimodal models.

BACKGROUND: Advances in artificial intelligence (AI) have realized the potential of revolutionizing healthcare, such as predicting disease progression via longitudinal inspection of Electronic Health Records (EHRs) and lab tests from patients admitted to Intensive Care Units (ICU). Although substantial literature exists addressing broad subjects, including the prediction of mortality, length-of-stay, and readmission, studies focusing on forecasting Acute Kidney Injury (AKI), specifically dialysis anticipation like Continuous Renal Replacement Therapy (CRRT) are scarce. The technicality of how to implement AI remains elusive. OBJECTIVE: This study aims to elucidate the important factors and methods that are required to develop effective predictive models of AKI and CRRT for patients admitted to ICU, using EHRs in the Medical Information Mart for Intensive Care (MIMIC) database. METHODS: We conducted a comprehensive comparative analysis of established predictive models, considering both time-series measurements and clinical notes from MIMIC-IV databases. Subsequently, we proposed a novel multi-modal model which integrates embeddings of top-performing unimodal models, including Long Short-Term Memory (LSTM) and BioMedBERT, and leverages both unstructured clinical notes and structured time series measurements derived from EHRs to enable the early prediction of AKI and CRRT. RESULTS: Our multimodal model achieved a lead time of at least 12 h ahead of clinical manifestation, with an Area Under the Receiver Operating Characteristic Curve (AUROC) of 0.888 for AKI and 0.997 for CRRT, as well as an Area Under the Precision Recall Curve (AUPRC) of 0.727 for AKI and 0.840 for CRRT, respectively, which significantly outperformed the baseline models. Additionally, we performed a SHapley Additive exPlanation (SHAP) analysis using the expected gradients algorithm, which highlighted important, previously underappreciated predictive features for AKI and CRRT. CONCLUSION: Our study revealed the importance and the technicality of applying longitudinal, multimodal modeling to improve early prediction of AKI and CRRT, offering insights for timely interventions. The performance and interpretability of our model indicate its potential for further assessment towards clinical applications, to ultimately optimize AKI management and enhance patient outcomes.

RNA2DNAlign: nucleotide resolution allele asymmetries through quantitative assessment of RNA and DNA paired sequencing data.

We introduce RNA2DNAlign, a computational framework for quantitative assessment of allele counts across paired RNA and DNA sequencing datasets. RNA2DNAlign is based on quantitation of the relative abundance of variant and reference read counts, followed by binomial tests for genotype and allelic status at SNV positions between compatible sequences. RNA2DNAlign detects positions with differential allele distribution, suggesting asymmetries due to regulatory/structural events. Based on the type of asymmetry, RNA2DNAlign outlines positions likely to be implicated in RNA editing, allele-specific expression or loss, somatic mutagenesis or loss-of-heterozygosity (the first three also in a tumor-specific setting). We applied RNA2DNAlign on 360 matching normal and tumor exomes and transcriptomes from 90 breast cancer patients from TCGA. Under high-confidence settings, RNA2DNAlign identified 2038 distinct SNV sites associated with one of the aforementioned asymetries, the majority of which have not been linked to functionality before. The performance assessment shows very high specificity and sensitivity, due to the corroboration of signals across multiple matching datasets. RNA2DNAlign is freely available from http://github.com/HorvathLab/NGS as a self-contained binary package for 64-bit Linux systems.

Forecasting Acute Kidney Injury and Resource Utilization in ICU patients using longitudinal, multimodal models.

Background: Advances in artificial intelligence (AI) have realized the potential of revolutionizing healthcare, such as predicting disease progression via longitudinal inspection of Electronic Health Records (EHRs) and lab tests from patients admitted to Intensive Care Units (ICU). Although substantial literature exists addressing broad subjects, including the prediction of mortality, length-of-stay, and readmission, studies focusing on forecasting Acute Kidney Injury (AKI), specifically dialysis anticipation like Continuous Renal Replacement Therapy (CRRT) are scarce. The technicality of how to implement AI remains elusive. Objective: This study aims to elucidate the important factors and methods that are required to develop effective predictive models of AKI and CRRT for patients admitted to ICU, using EHRs in the Medical Information Mart for Intensive Care (MIMIC) database. Methods: We conducted a comprehensive comparative analysis of established predictive models, considering both time-series measurements and clinical notes from MIMIC-IV databases. Subsequently, we proposed a novel multi-modal model which integrates embeddings of top-performing unimodal models, including Long Short-Term Memory (LSTM) and BioMedBERT, and leverages both unstructured clinical notes and structured time series measurements derived from EHRs to enable the early prediction of AKI and CRRT. Results: Our multimodal model achieved a lead time of at least 12 hours ahead of clinical manifestation, with an Area Under the Receiver Operating Characteristic Curve (AUROC) of 0.888 for AKI and 0.997 for CRRT, as well as an Area Under the Precision Recall Curve (AUPRC) of 0.727 for AKI and 0.840 for CRRT, respectively, which significantly outperformed the baseline models. Additionally, we performed a SHapley Additive exPlanation (SHAP) analysis using the expected gradients algorithm, which highlighted important, previously underappreciated predictive features for AKI and CRRT. Conclusion: Our study revealed the importance and the technicality of applying longitudinal, multimodal modeling to improve early prediction of AKI and CRRT, offering insights for timely interventions. The performance and interpretability of our model indicate its potential for further assessment towards clinical applications, to ultimately optimize AKI management and enhance patient outcomes.

Elucidating immune-related gene transcriptional programs via factorization of large-scale RNA-profiles.

Recent developments in immunotherapy, including immune checkpoint blockade (ICB) and adoptive cell therapy (ACT), have encountered challenges such as immune-related adverse events and resistance, especially in solid tumors. To advance the field, a deeper understanding of the molecular mechanisms behind treatment responses and resistance is essential. However, the lack of functionally characterized immune-related gene sets has limited data-driven immunological research. To address this gap, we adopted non-negative matrix factorization on 83 human bulk RNA sequencing (RNA-seq) datasets and constructed 28 immune-specific gene sets. After rigorous immunologist-led manual annotations and orthogonal validations across immunological contexts and functional omics data, we demonstrated that these gene sets can be applied to refine pan-cancer immune subtypes, improve ICB response prediction and functionally annotate spatial transcriptomic data. These functional gene sets, informing diverse immune states, will advance our understanding of immunology and cancer research.

Elucidating immune-related gene transcriptional programs via factorization of large-scale RNA-profiles.

Recent developments in immunotherapy, including immune checkpoint blockade (ICB) and adoptive cell therapy, have encountered challenges such as immune-related adverse events and resistance, especially in solid tumors. To advance the field, a deeper understanding of the molecular mechanisms behind treatment responses and resistance is essential. However, the lack of functionally characterized immune-related gene sets has limited data-driven immunological research. To address this gap, we adopted non-negative matrix factorization on 83 human bulk RNA-seq datasets and constructed 28 immune-specific gene sets. After rigorous immunologist-led manual annotations and orthogonal validations across immunological contexts and functional omics data, we demonstrated that these gene sets can be applied to refine pan-cancer immune subtypes, improve ICB response prediction and functionally annotate spatial transcriptomic data. These functional gene sets, informing diverse immune states, will advance our understanding of immunology and cancer research.

Interleukin-21 engineering enhances NK cell activity against glioblastoma via CEBPD.

Glioblastoma (GBM) is an aggressive brain cancer with limited therapeutic options. Natural killer (NK) cells are innate immune cells with strong anti-tumor activity and may offer a promising treatment strategy for GBM. We compared the anti-GBM activity of NK cells engineered to express interleukin (IL)-15 or IL-21. Using multiple in vivo models, IL-21 NK cells were superior to IL-15 NK cells both in terms of safety and long-term anti-tumor activity, with locoregionally administered IL-15 NK cells proving toxic and ineffective at tumor control. IL-21 NK cells displayed a unique chromatin accessibility signature, with CCAAT/enhancer-binding proteins (C/EBP), especially CEBPD, serving as key transcription factors regulating their enhanced function. Deletion of CEBPD resulted in loss of IL-21 NK cell potency while its overexpression increased NK cell long-term cytotoxicity and metabolic fitness. These results suggest that IL-21, through C/EBP transcription factors, drives epigenetic reprogramming of NK cells, enhancing their anti-tumor efficacy against GBM.

Safety, efficacy and determinants of response of allogeneic CD19-specific CAR-NK cells in CD19+ B cell tumors: a phase 1/2 trial.

There is a pressing need for allogeneic chimeric antigen receptor (CAR)-immune cell therapies that are safe, effective and affordable. We conducted a phase 1/2 trial of cord blood-derived natural killer (NK) cells expressing anti-CD19 chimeric antigen receptor and interleukin-15 (CAR19/IL-15) in 37 patients with CD19+ B cell malignancies. The primary objectives were safety and efficacy, defined as day 30 overall response (OR). Secondary objectives included day 100 response, progression-free survival, overall survival and CAR19/IL-15 NK cell persistence. No notable toxicities such as cytokine release syndrome, neurotoxicity or graft-versus-host disease were observed. The day 30 and day 100 OR rates were 48.6% for both. The 1-year overall survival and progression-free survival were 68% and 32%, respectively. Patients who achieved OR had higher levels and longer persistence of CAR-NK cells. Receiving CAR-NK cells from a cord blood unit (CBU) with nucleated red blood cells ≤ 8 × 107 and a collection-to-cryopreservation time ≤ 24 h was the most significant predictor for superior outcome. NK cells from these optimal CBUs were highly functional and enriched in effector-related genes. In contrast, NK cells from suboptimal CBUs had upregulation of inflammation, hypoxia and cellular stress programs. Finally, using multiple mouse models, we confirmed the superior antitumor activity of CAR/IL-15 NK cells from optimal CBUs in vivo. These findings uncover new features of CAR-NK cell biology and underscore the importance of donor selection for allogeneic cell therapies. ClinicalTrials.gov identifier: NCT03056339 .

Characterizing cancer metabolism from bulk and single-cell RNA-seq data using METAFlux.

Cells often alter metabolic strategies under nutrient-deprived conditions to support their survival and growth. Characterizing metabolic reprogramming in the tumor microenvironment (TME) is of emerging importance in cancer research and patient care. However, recent technologies only measure a subset of metabolites and cannot provide in situ measurements. Computational methods such as flux balance analysis (FBA) have been developed to estimate metabolic flux from bulk RNA-seq data and can potentially be extended to single-cell RNA-seq (scRNA-seq) data. However, it is unclear how reliable current methods are, particularly in TME characterization. Here, we present a computational framework METAFlux (METAbolic Flux balance analysis) to infer metabolic fluxes from bulk or single-cell transcriptomic data. Large-scale experiments using cell-lines, the cancer genome atlas (TCGA), and scRNA-seq data obtained from diverse cancer and immunotherapeutic contexts, including CAR-NK cell therapy, have validated METAFlux's capability to characterize metabolic heterogeneity and metabolic interaction amongst cell types.

Systematic Analysis of Actively Transcribed Core Matrisome Genes Across Tissues and Cell Phenotypes.

The extracellular matrix (ECM) is a highly dynamic, well-organized acellular network of tissue-specific biomolecules, that can be divided into structural or core ECM proteins and ECM-associated proteins. The ECM serves as a blueprint for organ development and function and, when structurally altered through mutation, altered expression, or degradation, can lead to debilitating syndromes that often affect one tissue more than another. Cross-referencing the FANTOM5 SSTAR (Semantic catalog of Samples, Transcription initiation And Regulators) and the defined catalog of core matrisome ECM (glyco)proteins, we conducted a comprehensive analysis of 511 different human samples to annotate the context-specific transcription of the individual components of the defined matrisome. Relative log expression normalized SSTAR cap analysis gene expression peak data files were downloaded from the FANTOM5 online database and filtered to exclude all cell lines and diseased tissues. Promoter-level expression values were categorized further into eight core tissue systems and three major ECM categories: proteoglycans, glycoproteins, and collagens. Hierarchical clustering and correlation analyses were conducted to identify complex relationships in promoter-driven gene expression activity. Integration of the core matrisome and curated FANTOM5 SSTAR data creates a unique tool that provides insight into the promoter-level expression of ECM-encoding genes in a tissue- and cell-specific manner. Unbiased clustering of cap analysis gene expression peak data reveals unique ECM signatures within defined tissue systems. Correlation analysis among tissue systems exposes both positive and negative correlation of ECM promoters with varying levels of significance. This tool can be used to provide new insight into the relationships between ECM components and tissues and can inform future research on the ECM in human disease and development. We invite the matrix biology community to continue to explore and discuss this dataset as part of a larger and continuing conversation about the human ECM. An interactive web tool can be found at matrixpromoterome.github.io along with additional resources that can be found at dx.doi.org/10.6084/m9.figshare.19794481 (figures) and https://figshare.com/s/e18ecbc3ae5aaf919b78 (python notebook).

Genome-wide CRISPR screens using isogenic cells reveal vulnerabilities conferred by loss of tumor suppressors.

Exploiting cancer vulnerabilities is critical for the discovery of anticancer drugs. However, tumor suppressors cannot be directly targeted because of their loss of function. To uncover specific vulnerabilities for cells with deficiency in any given tumor suppressor(s), we performed genome-scale CRISPR loss-of-function screens using a panel of isogenic knockout cells we generated for 12 common tumor suppressors. Here, we provide a comprehensive and comparative dataset for genetic interactions between the whole-genome protein-coding genes and a panel of tumor suppressor genes, which allows us to uncover known and new high-confidence synthetic lethal interactions. Mining this dataset, we uncover essential paralog gene pairs, which could be a common mechanism for interpreting synthetic lethality. Moreover, we propose that some tumor suppressors could be targeted to suppress proliferation of cells with deficiency in other tumor suppressors. This dataset provides valuable information that can be further exploited for targeted cancer therapy.

A Technique for Salvaging an Existing Four-Implant-Supported Mandibular Fixed Complete Prostheses by Using it as a Surgical Guide after an Implant Failure.

This technique article describes the use of an existing four implant supported fixed complete prosthesis both as a surgical template for reimplantation and for a permanent prosthesis after one of the supporting implants fails. This method offers a reliable and low-cost solution for the patient without the necessity of a new prosthesis.

Discovery of putative tumor suppressors from CRISPR screens reveals rewired lipid metabolism in acute myeloid leukemia cells.

CRISPR knockout fitness screens in cancer cell lines reveal many genes whose loss of function causes cell death or loss of fitness or, more rarely, the opposite phenotype of faster proliferation. Here we demonstrate a systematic approach to identify these proliferation suppressors, which are highly enriched for tumor suppressor genes, and define a network of 145 such genes in 22 modules. One module contains several elements of the glycerolipid biosynthesis pathway and operates exclusively in a subset of acute myeloid leukemia cell lines. The proliferation suppressor activity of genes involved in the synthesis of saturated fatty acids, coupled with a more severe loss of fitness phenotype for genes in the desaturation pathway, suggests that these cells operate at the limit of their carrying capacity for saturated fatty acids, which we confirm biochemically. Overexpression of this module is associated with a survival advantage in juvenile leukemias, suggesting a clinically relevant subtype.

Multiplex enCas12a screens detect functional buffering among paralogs otherwise masked in monogenic Cas9 knockout screens.

BACKGROUND: Pooled library CRISPR/Cas9 knockout screening across hundreds of cell lines has identified genes whose disruption leads to fitness defects, a critical step in identifying candidate cancer targets. However, the number of essential genes detected from these monogenic knockout screens is low compared to the number of constitutively expressed genes in a cell. RESULTS: Through a systematic analysis of screen data in cancer cell lines generated by the Cancer Dependency Map, we observe that half of all constitutively expressed genes are never detected in any CRISPR screen and that these never-essentials are highly enriched for paralogs. We investigated functional buffering among approximately 400 candidate paralog pairs using CRISPR/enCas12a dual-gene knockout screening in three cell lines. We observe 24 synthetic lethal paralog pairs that have escaped detection by monogenic knockout screens at stringent thresholds. Nineteen of 24 (79%) synthetic lethal interactions are present in at least two out of three cell lines and 14 of 24 (58%) are present in all three cell lines tested, including alternate subunits of stable protein complexes as well as functionally redundant enzymes. CONCLUSIONS: Together, these observations strongly suggest that functionally redundant paralogs represent a targetable set of genetic dependencies that are systematically under-represented among cell-essential genes in monogenic CRISPR-based loss of function screens.

A network of human functional gene interactions from knockout fitness screens in cancer cells.

Genetic interactions mediate the emergence of phenotype from genotype. The systematic survey of genetic interactions in yeast showed that genes operating in the same biological process have highly correlated genetic interaction profiles, and this observation has been exploited to infer gene function in model organisms. Such assays of digenic perturbations in human cells are also highly informative, but are not scalable, even with CRISPR-mediated methods. As an alternative, we developed an indirect method of deriving functional interactions. We show that genes having correlated knockout fitness profiles across diverse, non-isogenic cell lines are analogous to genes having correlated genetic interaction profiles across isogenic query strains and similarly imply shared biological function. We constructed a network of genes with correlated fitness profiles across 276 high-quality CRISPR knockout screens in cancer cell lines into a "coessentiality network," with up to 500-fold enrichment for co-functional gene pairs, enabling strong inference of gene function and highlighting the modular organization of the cell.