Deep Learning-based Propensity Scores for Confounding Control in Comparative Effectiveness Research: A Large-scale, Real-world Data Study.

BACKGROUND: Due to the non-randomized nature of real-world data, prognostic factors need to be balanced, which is often done by propensity scores (PSs). This study aimed to investigate whether autoencoders, which are unsupervised deep learning architectures, might be leveraged to compute PS. METHODS: We selected patient-level data of 128,368 first-line treated cancer patients from the Flatiron Health EHR-derived de-identified database. We trained an autoencoder architecture to learn a lower-dimensional patient representation, which we used to compute PS. To compare the performance of an autoencoder-based PS with established methods, we performed a simulation study. We assessed the balancing and adjustment performance using standardized mean differences, root mean square errors (RMSE), percent bias, and confidence interval coverage. To illustrate the application of the autoencoder-based PS, we emulated the PRONOUNCE trial by applying the trial's protocol elements within an observational database setting, comparing two chemotherapy regimens. RESULTS: All methods but the manual variable selection approach led to well-balanced cohorts with average standardized mean differences <0.1. LASSO yielded on average the lowest deviation of resulting estimates (RMSE 0.0205) followed by the autoencoder approach (RMSE 0.0248). Altering the hyperparameter setup in sensitivity analysis, the autoencoder approach led to similar results as LASSO (RMSE 0.0203 and 0.0205, respectively). In the case study, all methods provided a similar conclusion with point estimates clustered around the null (e.g., HRautoencoder 1.01 [95% confidence interval = 0.80, 1.27] vs. HRPRONOUNCE 1.07 [0.83, 1.36]). CONCLUSIONS: Autoencoder-based PS computation was a feasible approach to control for confounding but did not perform better than some established approaches like LASSO.

netprioR: a probabilistic model for integrative hit prioritisation of genetic screens.

In the post-genomic era of big data in biology, computational approaches to integrate multiple heterogeneous data sets become increasingly important. Despite the availability of large amounts of omics data, the prioritisation of genes relevant for a specific functional pathway based on genetic screening experiments, remains a challenging task. Here, we introduce netprioR, a probabilistic generative model for semi-supervised integrative prioritisation of hit genes. The model integrates multiple network data sets representing gene-gene similarities and prior knowledge about gene functions from the literature with gene-based covariates, such as phenotypes measured in genetic perturbation screens, for example, by RNA interference or CRISPR/Cas9. We evaluate netprioR on simulated data and show that the model outperforms current state-of-the-art methods in many scenarios and is on par otherwise. In an application to real biological data, we integrate 22 network data sets, 1784 prior knowledge class labels and 3840 RNA interference phenotypes in order to prioritise novel regulators of Notch signalling in Drosophila melanogaster. The biological relevance of our predictions is evaluated using in silico and in vivo experiments. An efficient implementation of netprioR is available as an R package at http://bioconductor.org/packages/netprioR.

Improved pathway reconstruction from RNA interference screens by exploiting off-target effects.

Motivation: Pathway reconstruction has proven to be an indispensable tool for analyzing the molecular mechanisms of signal transduction underlying cell function. Nested effects models (NEMs) are a class of probabilistic graphical models designed to reconstruct signalling pathways from high-dimensional observations resulting from perturbation experiments, such as RNA interference (RNAi). NEMs assume that the short interfering RNAs (siRNAs) designed to knockdown specific genes are always on-target. However, it has been shown that most siRNAs exhibit strong off-target effects, which further confound the data, resulting in unreliable reconstruction of networks by NEMs. Results: Here, we present an extension of NEMs called probabilistic combinatorial nested effects models (pc-NEMs), which capitalize on the ancillary siRNA off-target effects for network reconstruction from combinatorial gene knockdown data. Our model employs an adaptive simulated annealing search algorithm for simultaneous inference of network structure and error rates inherent to the data. Evaluation of pc-NEMs on simulated data with varying number of phenotypic effects and noise levels as well as real data demonstrates improved reconstruction compared to classical NEMs. Application to Bartonella henselae infection RNAi screening data yielded an eight node network largely in agreement with previous works, and revealed novel binary interactions of direct impact between established components. Availability and implementation: The software used for the analysis is freely available as an R package at https://github.com/cbg-ethz/pcNEM.git. Supplementary information: Supplementary data are available at Bioinformatics online.

Properties of short double-stranded RNAs carrying randomized base pairs: toward better controls for RNAi experiments.

Short interfering RNAs (siRNAs) are mediators of RNA interference (RNAi), a commonly used technique for selective down-regulation of target gene expression. Using an equimolar mixture of A, G, C, and U phosphoramidites during solid-phase synthesis, we introduced degenerate positions into RNA guide and passenger strands so that, when annealed, a large pool of distinct siRNA duplexes with randomized base pairs at defined sites was created. We assessed the randomization efficiency by deep sequencing one of the RNAs. All possible individual sequences were present in the pool with generally an excellent distribution of bases. Melting temperature analyses suggested that pools of randomized guide and passenger strands RNAs with up to eight degenerate positions annealed so that mismatched base-pairing was minimized. Transfections of randomized siRNAs (rnd-siRNAs) into cells led to inhibition of luciferase reporters by a miRNA-like mechanism when the seed regions of rnd-siRNA guide strands were devoid of degenerate positions. Furthermore, the mRNA levels of a select set of genes associated with siRNA off-target effects were measured and indicated that rnd-siRNAs with degenerate positions in the seed likely show typical non-sequence-specific effects, but not miRNA-like off-target effects. In the wake of recent reports showing the preponderance of miRNA-like off-target effects of siRNAs, our findings are of value for the design of a novel class of easily prepared and universally applicable negative siRNA controls.

Identification of hypoxia-regulated proteins using MALDI-mass spectrometry imaging combined with quantitative proteomics.

Hypoxia is present in most solid tumors and is clinically correlated with increased metastasis and poor patient survival. While studies have demonstrated the role of hypoxia and hypoxia-regulated proteins in cancer progression, no attempts have been made to identify hypoxia-regulated proteins using quantitative proteomics combined with MALDI-mass spectrometry imaging (MALDI-MSI). Here we present a comprehensive hypoxic proteome study and are the first to investigate changes in situ using tumor samples. In vitro quantitative mass spectrometry analysis of the hypoxic proteome was performed on breast cancer cells using stable isotope labeling with amino acids in cell culture (SILAC). MS analyses were performed on laser-capture microdissected samples isolated from normoxic and hypoxic regions from tumors derived from the same cells used in vitro. MALDI-MSI was used in combination to investigate hypoxia-regulated protein localization within tumor sections. Here we identified more than 100 proteins, both novel and previously reported, that were associated with hypoxia. Several proteins were localized in hypoxic regions, as identified by MALDI-MSI. Visualization and data extrapolation methods for the in vitro SILAC data were also developed, and computational mapping of MALDI-MSI data to IHC results was applied for data validation. The results and limitations of the methodologies described are discussed.

Simultaneous analysis of large-scale RNAi screens for pathogen entry.

BACKGROUND: Large-scale RNAi screening has become an important technology for identifying genes involved in biological processes of interest. However, the quality of large-scale RNAi screening is often deteriorated by off-targets effects. In order to find statistically significant effector genes for pathogen entry, we systematically analyzed entry pathways in human host cells for eight pathogens using image-based kinome-wide siRNA screens with siRNAs from three vendors. We propose a Parallel Mixed Model (PMM) approach that simultaneously analyzes several non-identical screens performed with the same RNAi libraries. RESULTS: We show that PMM gains statistical power for hit detection due to parallel screening. PMM allows incorporating siRNA weights that can be assigned according to available information on RNAi quality. Moreover, PMM is able to estimate a sharedness score that can be used to focus follow-up efforts on generic or specific gene regulators. By fitting a PMM model to our data, we found several novel hit genes for most of the pathogens studied. CONCLUSIONS: Our results show parallel RNAi screening can improve the results of individual screens. This is currently particularly interesting when large-scale parallel datasets are becoming more and more publicly available. Our comprehensive siRNA dataset provides a public, freely available resource for further statistical and biological analyses in the high-content, high-throughput siRNA screening field.

Generative artificial intelligence empowers digital twins in drug discovery and clinical trials.

INTRODUCTION: The concept of Digital Twins (DTs) translated to drug development and clinical trials describes virtual representations of systems of various complexities, ranging from individual cells to entire humans, and enables in silico simulations and experiments. DTs increase the efficiency of drug discovery and development by digitalizing processes associated with high economic, ethical, or social burden. The impact is multifaceted: DT models sharpen disease understanding, support biomarker discovery and accelerate drug development, thus advancing precision medicine. One way to realize DTs is by generative artificial intelligence (AI), a cutting-edge technology that enables the creation of novel, realistic and complex data with desired properties. AREAS COVERED: The authors provide a brief introduction to generative AI and describe how it facilitates the modeling of DTs. In addition, they compare existing implementations of generative AI for DTs in drug discovery and clinical trials. Finally, they discuss technical and regulatory challenges that should be addressed before DTs can transform drug discovery and clinical trials. EXPERT OPINION: The current state of DTs in drug discovery and clinical trials does not exploit the entire power of generative AI yet and is limited to simulation of a small number of characteristics. Nonetheless, generative AI has the potential to transform the field by leveraging recent developments in deep learning and customizing models for the needs of scientists, physicians and patients.

PXPermute reveals staining importance in multichannel imaging flow cytometry.

Imaging flow cytometry (IFC) allows rapid acquisition of numerous single-cell images per second, capturing information from multiple fluorescent channels. However, the traditional process of staining cells with fluorescently labeled conjugated antibodies for IFC analysis is time consuming, expensive, and potentially harmful to cell viability. To streamline experimental workflows and reduce costs, it is crucial to identify the most relevant channels for downstream analysis. In this study, we introduce PXPermute, a user-friendly and powerful method for assessing the significance of IFC channels, particularly for cell profiling. Our approach evaluates channel importance by permuting pixel values within each channel and analyzing the resulting impact on machine learning or deep learning models. Through rigorous evaluation of three multichannel IFC image datasets, we demonstrate PXPermute's potential in accurately identifying the most informative channels, aligning with established biological knowledge. PXPermute can assist biologists with systematic channel analysis, experimental design optimization, and biomarker identification.

Explainable machine learning for profiling the immunological synapse and functional characterization of therapeutic antibodies.

Therapeutic antibodies are widely used to treat severe diseases. Most of them alter immune cells and act within the immunological synapse; an essential cell-to-cell interaction to direct the humoral immune response. Although many antibody designs are generated and evaluated, a high-throughput tool for systematic antibody characterization and prediction of function is lacking. Here, we introduce the first comprehensive open-source framework, scifAI (single-cell imaging flow cytometry AI), for preprocessing, feature engineering, and explainable, predictive machine learning on imaging flow cytometry (IFC) data. Additionally, we generate the largest publicly available IFC dataset of the human immunological synapse containing over 2.8 million images. Using scifAI, we analyze class frequency and morphological changes under different immune stimulation. T cell cytokine production across multiple donors and therapeutic antibodies is quantitatively predicted in vitro, linking morphological features with function and demonstrating the potential to significantly impact antibody design. scifAI is universally applicable to IFC data. Given its modular architecture, it is straightforward to incorporate into existing workflows and analysis pipelines, e.g., for rapid antibody screening and functional characterization.

Computational Tumor Infiltration Phenotypes Enable the Spatial and Genomic Analysis of Immune Infiltration in Colorectal Cancer.

Cancer immunotherapy has led to significant therapeutic progress in the treatment of metastatic and formerly untreatable tumors. However, drug response rates are variable and often only a subgroup of patients will show durable response to a treatment. Biomarkers that help to select those patients that will benefit the most from immunotherapy are thus of crucial importance. Here, we aim to identify such biomarkers by investigating the tumor microenvironment, i.e., the interplay between different cell types like immune cells, stromal cells and malignant cells within the tumor and developed a computational method that determines spatial tumor infiltration phenotypes. Our method is based on spatial point pattern analysis of immunohistochemically stained colorectal cancer tumor tissue and accounts for the intra-tumor heterogeneity of immune infiltration. We show that, compared to base-line models, tumor infiltration phenotypes provide significant additional support for the prediction of established biomarkers in a colorectal cancer patient cohort (n = 80). Integration of tumor infiltration phenotypes with genetic and genomic data from the same patients furthermore revealed significant associations between spatial infiltration patterns and common mutations in colorectal cancer and gene expression signatures. Based on these associations, we computed novel gene signatures that allow one to predict spatial tumor infiltration patterns from gene expression data only and validated this approach in a separate dataset from the Cancer Genome Atlas.

Learning epistatic gene interactions from perturbation screens.

The treatment of complex diseases often relies on combinatorial therapy, a strategy where drugs are used to target multiple genes simultaneously. Promising candidate genes for combinatorial perturbation often constitute epistatic genes, i.e., genes which contribute to a phenotype in a non-linear fashion. Experimental identification of the full landscape of genetic interactions by perturbing all gene combinations is prohibitive due to the exponential growth of testable hypotheses. Here we present a model for the inference of pairwise epistatic, including synthetic lethal, gene interactions from siRNA-based perturbation screens. The model exploits the combinatorial nature of siRNA-based screens resulting from the high numbers of sequence-dependent off-target effects, where each siRNA apart from its intended target knocks down hundreds of additional genes. We show that conditional and marginal epistasis can be estimated as interaction coefficients of regression models on perturbation data. We compare two methods, namely glinternet and xyz, for selecting non-zero effects in high dimensions as components of the model, and make recommendations for the appropriate use of each. For data simulated from real RNAi screening libraries, we show that glinternet successfully identifies epistatic gene pairs with high accuracy across a wide range of relevant parameters for the signal-to-noise ratio of observed phenotypes, the effect size of epistasis and the number of observations per double knockdown. xyz is also able to identify interactions from lower dimensional data sets (fewer genes), but is less accurate for many dimensions. Higher accuracy of glinternet, however, comes at the cost of longer running time compared to xyz. The general model is widely applicable and allows mining the wealth of publicly available RNAi screening data for the estimation of epistatic interactions between genes. As a proof of concept, we apply the model to search for interactions, and potential targets for treatment, among previously published sets of siRNA perturbation screens on various pathogens. The identified interactions include both known epistatic interactions as well as novel findings.

Context-dependent deposition and regulation of mRNAs in P-bodies.

Cells respond to stress by remodeling their transcriptome through transcription and degradation. Xrn1p-dependent degradation in P-bodies is the most prevalent decay pathway, yet, P-bodies may facilitate not only decay, but also act as a storage compartment. However, which and how mRNAs are selected into different degradation pathways and what determines the fate of any given mRNA in P-bodies remain largely unknown. We devised a new method to identify both common and stress-specific mRNA subsets associated with P-bodies. mRNAs targeted for degradation to P-bodies, decayed with different kinetics. Moreover, the localization of a specific set of mRNAs to P-bodies under glucose deprivation was obligatory to prevent decay. Depending on its client mRNA, the RNA-binding protein Puf5p either promoted or inhibited decay. Furthermore, the Puf5p-dependent storage of a subset of mRNAs in P-bodies under glucose starvation may be beneficial with respect to chronological lifespan.

A Genome-Wide siRNA Screen Implicates Spire1/2 in SipA-Driven Salmonella Typhimurium Host Cell Invasion.

Salmonella Typhimurium (S. Tm) is a leading cause of diarrhea. The disease is triggered by pathogen invasion into the gut epithelium. Invasion is attributed to the SPI-1 type 3 secretion system (T1). T1 injects effector proteins into epithelial cells and thereby elicits rearrangements of the host cellular actin cytoskeleton and pathogen invasion. The T1 effector proteins SopE, SopB, SopE2 and SipA are contributing to this. However, the host cell factors contributing to invasion are still not completely understood. To address this question comprehensively, we used Hela tissue culture cells, a genome-wide siRNA library, a modified gentamicin protection assay and S. TmSipA, a sopBsopE2sopE mutant which strongly relies on the T1 effector protein SipA to invade host cells. We found that S. TmSipA invasion does not elicit membrane ruffles, nor promote the entry of non-invasive bacteria "in trans". However, SipA-mediated infection involved the SPIRE family of actin nucleators, besides well-established host cell factors (WRC, ARP2/3, RhoGTPases, COPI). Stage-specific follow-up assays and knockout fibroblasts indicated that SPIRE1 and SPIRE2 are involved in different steps of the S. Tm infection process. Whereas SPIRE1 interferes with bacterial binding, SPIRE2 influences intracellular replication of S. Tm. Hence, these two proteins might fulfill non-redundant functions in the pathogen-host interaction. The lack of co-localization hints to a short, direct interaction between S. Tm and SPIRE proteins or to an indirect effect.

gespeR: a statistical model for deconvoluting off-target-confounded RNA interference screens.

Small interfering RNAs (siRNAs) exhibit strong off-target effects, which confound the gene-level interpretation of RNA interference screens and thus limit their utility for functional genomics studies. Here, we present gespeR, a statistical model for reconstructing individual, gene-specific phenotypes. Using 115,878 siRNAs, single and pooled, from three companies in three pathogen infection screens, we demonstrate that deconvolution of image-based phenotypes substantially improves the reproducibility between independent siRNA sets targeting the same genes. Genes selected and prioritized by gespeR are validated and shown to constitute biologically relevant components of pathogen entry mechanisms and TGF-ß signaling. gespeR is available as a Bioconductor R-package.

Global characterization of signalling networks associated with tamoxifen resistance in breast cancer.

Acquired resistance to the anti-estrogen tamoxifen remains a significant challenge in breast cancer management. In this study, we used an integrative approach to characterize global protein expression and tyrosine phosphorylation events in tamoxifen-resistant MCF7 breast cancer cells (TamR) compared with parental controls. Quantitative mass spectrometry and computational approaches were combined to identify perturbed signalling networks, and candidate regulatory proteins were functionally interrogated by siRNA-mediated knockdown. Network analysis revealed that cellular metabolism was perturbed in TamR cells, together with pathways enriched for proteins associated with growth factor, cell-cell and cell matrix-initiated signalling. Consistent with known roles for Ras/MAPK and PI3-kinase signalling in tamoxifen resistance, tyrosine-phosphorylated MAPK1, SHC1 and PIK3R2 were elevated in TamR cells. Phosphorylation of the tyrosine kinase Yes and expression of the actin-binding protein myristoylated alanine-rich C-kinase substrate (MARCKS) were increased two- and eightfold in TamR cells respectively, and these proteins were selected for further analysis. Knockdown of either protein in TamR cells had no effect on anti-estrogen sensitivity, but significantly decreased cell motility. MARCKS expression was significantly higher in breast cancer cell lines than normal mammary epithelial cells and in ER-negative versus ER-positive breast cancer cell lines. In primary breast cancers, cytoplasmic MARCKS staining was significantly higher in basal-like and HER2 cancers than in luminal cancers, and was independently predictive of poor survival in multivariate analyses of the whole cohort (P < 0.0001) and in ER-positive patients (P = 0.0005). These findings provide network-level insights into the molecular alterations associated with the tamoxifen-resistant phenotype, and identify MARCKS as a potential biomarker of therapeutic responsiveness that may assist in stratification of patients for optimal therapy.