The TRANSFAC® database on transcription factors, their binding sites, nucleotide distribution matrices and regulated genes as well as the complementing database TRANSCompel® on composite elements have been further enhanced on various levels. A new web interface with different search options and integrated versions of Match™ and Patch™ provides increased functionality for TRANSFAC®. The list of databases which are linked to the common GENE table of TRANSFAC® and TRANSCompel® has been extended by: Ensembl, UniGene, EntrezGene, HumanPSD™ and TRANSPRO™. Standard gene names from HGNC, MGI and RGD, are included for human, mouse and rat genes, respectively. With the help of InterProScan, Pfam, SMART and PROSITE domains are assigned automatically to the protein sequences of the transcription factors. TRANSCompel® contains now, in addition to the COMPEL table, a separate table for detailed information on the experimental EVIDENCE on which the composite elements are based. Finally, for TRANSFAC®, in respect of data growth, in particular the gain of Drosophila transcription factor binding sites (by courtesy of the Drosophila DNase I footprint database) and of Arabidopsis factors (by courtesy of DATF, Database of Arabidopsis Transcription Factors) has to be stressed. The here described public releases, TRANSFAC® 7.0 and TRANSCompel® 7.0, are accessible under .

Background Contemporary deep learning approaches show cutting-edge performance in a variety of complex prediction tasks. Nonetheless, the application of deep learning in healthcare remains limited since deep learning methods are often considered as non-interpretable black-box models. However, the machine learning community made recent elaborations on interpretability methods explaining data point-specific decisions of deep learning techniques. We believe that such explanations can assist the need in personalized precision medicine decisions via explaining patient-specific predictions. Methods Layer-wise Relevance Propagation (LRP) is a technique to explain decisions of deep learning methods. It is widely used to interpret Convolutional Neural Networks (CNNs) applied on image data. Recently, CNNs started to extend towards non-Euclidean domains like graphs. Molecular networks are commonly represented as graphs detailing interactions between molecules. Gene expression data can be assigned to the vertices of these graphs. In other words, gene expression data can be structured by utilizing molecular network information as prior knowledge. Graph-CNNs can be applied to structured gene expression data, for example, to predict metastatic events in breast cancer. Therefore, there is a need for explanations showing which part of a molecular network is relevant for predicting an event, e.g., distant metastasis in cancer, for each individual patient. Results We extended the procedure of LRP to make it available for Graph-CNN and tested its applicability on a large breast cancer dataset. We present Graph Layer-wise Relevance Propagation (GLRP) as a new method to explain the decisions made by Graph-CNNs. We demonstrate a sanity check of the developed GLRP on a hand-written digits dataset and then apply the method on gene expression data. We show that GLRP provides patient-specific molecular subnetworks that largely agree with clinical knowledge and identify common as well as novel, and potentially druggable, drivers of tumor progression. Conclusions The developed method could be potentially highly useful on interpreting classification results in the context of different omics data and prior knowledge molecular networks on the individual patient level, as for example in precision medicine approaches or a molecular tumor board.

A strategy is presented that allows a causal analysis of co-expressed genes, which may be subject to common regulatory influences. A state-of-the-art promoter analysis for potential transcription factor (TF) binding sites in combination with a knowledge-based analysis of the upstream pathway that control the activity of these TFs is shown to lead to hypothetical master regulators. This strategy was implemented as a workflow in a comprehensive bioinformatic software platform. We applied this workflow to gene sets that were identified by a novel triclustering algorithm in naphthalene-induced gene expression signatures of murine liver and lung tissue. As a result, tissue-specific master regulators were identified that are known to be linked with tumorigenic and apoptotic processes. To our knowledge, this is the first time that genes of expression triclusters were used to identify upstream regulators.

Composite Module Analyst (CMA) is a novel software tool aiming to identify promoter-enhancer models based on the composition of transcription factor (TF) binding sites and their pairs. CMA is closely interconnected with the TRANSFAC® database. In particular, CMA uses the positional weight matrix (PWM) library collected in TRANSFAC® and therefore provides the possibility to search for a large variety of different TF binding sites. We model the structure of the long gene regulatory regions by a Boolean function that joins several local modules, each consisting of co-localized TF binding sites. Having as an input a set of co-regulated genes, CMA builds the promoter model and optimizes the parameters of the model automatically by applying a genetic-regression algorithm. We use a multicomponent fitness function of the algorithm which includes several statistical criteria in a weighted linear function. We show examples of successful application of CMA to a microarray data on transcription profiling of TNF-alpha stimulated primary human endothelial cells. The CMA web server is freely accessible at . An advanced version of CMA is also a part of the commercial system ExPlain™ () designed for causal analysis of gene expression data.

Graphical abstract