MOTIVATION: The exponential growth of genomic sequencing data has created ever-expanding repositories of gene networks. Unsupervised network integration methods are critical to learn informative representations for each gene, which are later used as features for downstream applications. However, these network integration methods must be scalable to account for the increasing number of networks and robust to an uneven distribution of network types within hundreds of gene networks. RESULTS: To address these needs, we present Gemini, a novel network integration method that uses memory-efficient high-order pooling to represent and weight each network according to its uniqueness. Gemini then mitigates the uneven network distribution through mixing up existing networks to create many new networks. We find that Gemini leads to more than a 10% improvement in F1 score, 15% improvement in micro-AUPRC, and 63% improvement in macro-AUPRC for human protein function prediction by integrating hundreds of networks from BioGRID, and that Gemini's performance significantly improves when more networks are added to the input network collection, while Mashup and BIONIC embeddings' performance deteriorates. Gemini thereby enables memory-efficient and informative network integration for large gene networks and can be used to massively integrate and analyze networks in other domains. AVAILABILITY AND IMPLEMENTATION: Gemini can be accessed at: https://github.com/MinxZ/Gemini.

Multi-modal learning is essential for understanding information in the real world. Jointly learning from multi-modal data enables global integration of both shared and modality-specific information, but current strategies often fail when observa- tions from certain modalities are incomplete or missing for part of the subjects. To learn comprehensive representations based on such modality-incomplete data, we present a semi-supervised neural network model called CLUE (Cross-Linked Unified Embedding). Extending from multi-modal VAEs, CLUE introduces the use of cross-encoders to construct latent representations from modality-incomplete observations. Representation learning for modality-incomplete observations is common in genomics. For example, human cells are tightly regulated across multi- ple related but distinct modalities such as DNA, RNA, and protein, jointly defining a cell’s function. We benchmark CLUE on multi-modal data from single cell measurements, illustrating CLUE’s superior performance in all assessed categories of the NeurIPS 2021 Multimodal Single-cell Data Integration Competition. While we focus on analysis of single cell genomic datasets, we note that the proposed cross-linked embedding strategy could be readily applied to other cross-modality representation learning problems.

MOTIVATION: High-dimensional DNA methylation (DNAm) array coverage, while sparse in the context of the entire DNA methylome, still constitutes a very large number of CpG probes. The ensuing multiple-test corrections affect the statistical power to detect associations, likely contributing to prevalent limited reproducibility. Array probes measuring proximal CpG sites often have correlated levels of DNAm that may not only be biologically meaningful but also imply statistical dependence and redundancy. New methods that account for such correlations between adjacent probes may enable improved specificity, discovery and interpretation of statistical associations in DNAm array data. RESULTS: We developed a method named Co-Methylation with genomic CpG Background (CoMeBack) that estimates DNA co-methylation, defined as proximal CpG probes with correlated DNAm across individuals. CoMeBack outputs co-methylated regions (CMRs), spanning sets of array probes constructed based on all genomic CpG sites, including those not measured on the array, and without any phenotypic variable inputs. This approach can reduce the multiple-test correction burden, while enhancing the discovery and specificity of statistical associations. We constructed and validated CMRs in whole blood, using publicly available Illumina Infinium 450 K array data from over 5000 individuals. These CMRs were enriched for enhancer chromatin states, and binding site motifs for several transcription factors involved in blood physiology. We illustrated how CMR-based epigenome-wide association studies can improve discovery and reduce false positives for associations with chronological age. AVAILABILITY AND IMPLEMENTATION: https://bitbucket.org/flopflip/comeback. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Complexity of cell-type composition has created much skepticism surrounding the interpretation of bulk tissue transcriptomic studies. Recent studies have shown that deconvolution algorithms can be applied to computationally estimate cell-type proportions from gene expression data of bulk blood samples, but their performance when applied to brain tissue is unclear. Here, we have generated an immunohistochemistry (IHC) dataset for five major cell-types from brain tissue of 70 individuals, who also have bulk cortical gene expression data. With the IHC data as the benchmark, this resource enables quantitative assessment of deconvolution algorithms for brain tissue. We apply existing deconvolution algorithms to brain tissue by using marker sets derived from human brain single cell and cell-sorted RNA-seq data. We show that these algorithms can indeed produce informative estimates of constituent cell-type proportions. In fact, neuronal subpopulations can also be estimated from bulk brain tissue samples. Further, we show that including the cell-type proportion estimates as confounding factors is important for reducing false associations between Alzheimer's disease phenotypes and gene expression. Lastly, we demonstrate that using more accurate marker sets can substantially improve statistical power in detecting cell-type specific expression quantitative trait loci (eQTLs).

To understand the regulation of tissue-specific gene expression, the GTEx Consortium generated RNA-seq expression data for more than thirty distinct human tissues. This data provides an opportunity for deriving shared and tissue specific gene regulatory networks on the basis of co-expression between genes. However, a small number of samples are available for a majority of the tissues, and therefore statistical inference of networks in this setting is highly underpowered. To address this problem, we infer tissue-specific gene co-expression networks for 35 tissues in the GTEx dataset using a novel algorithm, GNAT, that uses a hierarchy of tissues to share data between related tissues. We show that this transfer learning approach increases the accuracy with which networks are learned. Analysis of these networks reveals that tissue-specific transcription factors are hubs that preferentially connect to genes with tissue specific functions. Additionally, we observe that genes with tissue-specific functions lie at the peripheries of our networks. We identify numerous modules enriched for Gene Ontology functions, and show that modules conserved across tissues are especially likely to have functions common to all tissues, while modules that are upregulated in a particular tissue are often instrumental to tissue-specific function. Finally, we provide a web tool, available at mostafavilab.stat.ubc.ca/GNAT, which allows exploration of gene function and regulation in a tissue-specific manner.

MOTIVATION: Lung cancer is often discovered long after its onset, making identifying genes important in its initiation and progression a challenge. By the time the tumors are discovered, we only observe the final sum of changes of the few genes that initiated cancer and thousands of genes that they have influenced. Gene interactions and heterogeneity of samples make it difficult to identify genes consistent between different cohorts. Using gene and gene-product interaction networks, we propose a principled approach to identify a small subset of genes whose network neighbors exhibit consistently high expression change (in cancerous tissue versus normal) regardless of their own expression. We hypothesize that these genes can shed light on the larger scale perturbations in the overall landscape of expression levels. RESULTS: We benchmark our method on simulated data, and show that we can recover a true gene list in noisy measurement data. We then apply our method to four non-small cell lung cancer and two pancreatic cancer cohorts, finding several genes that are consistent within all cohorts of the same cancer type. CONCLUSION: Our model is flexible, robust and identifies gene sets that are more consistent across cohorts than several other approaches. Additionally, our method can be applied on a per-patient basis not requiring large cohorts of patients to find genes of influence. Our approach is generally applicable to gene expression studies where the goal is to identify a small set of influential genes that may in turn explain the much larger set of genome-wide expression changes.

MOTIVATION: Many algorithms that integrate multiple functional association networks for predicting gene function construct a composite network as a weighted sum of the individual networks and then use the composite network to predict gene function. The weight assigned to an individual network represents the usefulness of that network in predicting a given gene function. However, because many categories of gene function have a small number of annotations, the process of assigning these network weights is prone to overfitting. RESULTS: Here, we address this problem by proposing a novel approach to combining multiple functional association networks. In particular, we present a method where network weights are simultaneously optimized on sets of related function categories. The method is simpler and faster than existing approaches. Further, we show that it produces composite networks with improved function prediction accuracy using five example species (yeast, mouse, fly, Esherichia coli and human). AVAILABILITY: Networks and code are available from: http://morrislab.med.utoronto.ca/sara/SW

Recombinant interferon beta (IFNβ) is one of the most commonly prescribed treatments for multiple sclerosis; however, the treatment results in partial success producing no benefit in almost half of the patients. We address the problem of identifying minimal and robust sets of molecular biomarkers that are able to present predictive models of response to treatment in multiple sclerosis patients. To achieve this, we utilize a multivariate feature selection and classification framework; OSeMA (orthogonal search model analysis) integrates fast orthogonal search algorithm for feature selection and discriminant analysis for classification. Feature-selection and classification performance of OSeMA are evaluated through comparative studies with two wrapper-approach feature-selection/classification systems. It is demonstrated that the feature-selection of OSeMA significantly reduces the computational time of exhaustive searches while identifying complex gene-gene relationships. Utilizing OSeMA, we are able to construct classification models that are highly predictive of therapy response in MS patients, based on their gene expression data acquired prior to initiation of IFNp treatment