Software and platforms hosted by our lab

Systems Medicine - Network-enhanced biomedical decision making

GrandForest is an online tool for de novo endophenotyping and mechanotyping of diseases using omics data. PathClass is an online tool for breast cancer subtyping from gene expression data. Instead of using gene panels, we employ pathway activity patterns for this purpose, which render statistically way more robust. Our web application allows to predict breast cancer subtypes for new samples, including custom uploaded samples as well as samples found on the Gene Expression Omnibus. Furthermore, selected pathways used for the individual predictors can be investigated.

Network enrichment

KeyPathwayMiner is a software for de novo network enrichment, aka network modules. It combines multiple OMICS data sets with biological networks to turn your expression, mutation, or association study into a systems biology story. It comes as Cytoscape app, as standalone software and as web service. It was downloaded >5000x from the Cytoscape app store alone and applied in various different biomedical settings.

 

TiCoNE is a software for the analysis of time course expression data (e.g. gene expression) together with biological networks. It will find time patterns emerging in the expression data and check for network modules enriched with genes of similar expression behavior over time. It comes as web server and as Cytoscape plugin. 

Competing endogenous RNA networks inference

Genes carry binding sites that allows specific microRNAs to repress their expression. According to the competing endogenous RNA hypothesis, genes also regulate each other in a competitive fashion via sponging microRNAs. We have contributed to the development of the (partial) correlation-based method SPONGE and the (conditional) mutual information-based method JAMI to infer such microRNA-mediated gene-gene interactions. In particular SPONGE is fast enough to infer ceRNA interactions genome-wide and thus facilitates the inference of a ceRNA regulatory network that can be used for hypothesis generation, biomarker detection and drug target discovery.

EWAS data analysis - Detection of differentially methylated regions

DiMmer is a Java tool for efficient processing of Illumina 450K and 850K EPIC chip data. It is fully parallelizable and can process even big cohort data sets. With DiMmer, one may now even compute empirical p-values using permutation tests and find differentially methylated regions including different strategies for correction for multiple testing. It can correct for cell composition effects and other confounders. Continuous outcome variables and confounders can be directly integrated using regression models. It comes with an intuitive user interface that guides the clinician or biologist through all steps of a typical Illumina EWAS chip data analysis.

Disease-specific copy number variation (CNV) identification

CoNVaQ is a web service for CNV-based association study between two or more sample groups. The web interface allows one to quickly upload segmented CNV samples and search for variations that are overrepresented in a population. CoNVaQ differs from previous tools by allowing you to specify which CNVRs are considered significant using simple queries, e.g. “find the largest region duplicated in > 70% of cases and < 10% of controls”.

Location-specific analysis of microbial profiles

BioAtlas is a novel user-friendly platform to ease ecologists in accessing the power of microbial 16S rRNA profiling in a location-specific context. The BioAtlas web interface allows several kinds of interactive analyses: 1) Browsing the geographical distribution of microbial taxonomies across the planet in google maps, 2) across the human body, and 3) across user-defined maps and using taxonomic trees. 4) Browsing the distribution of the occurring taxonomies in user data against previously discovered

taxonomies across the planet, the human body, and/or own user maps. 6) Building own user maps (e.g. for plants, animals, …) online and mapping of taxonomies to the locations, and follow-up comparison of additional samples and taxonomy distributions with these maps.

High-throughput screening

HitSeekR is the first web platform for analyzing high-throughput screening (HTS) data of various types, from miRNA (inhibitor) screens and RNAi assays to CRISPER/cas9 and drug response screens. It can accommodate, normalize, etc. small to ultra-large scale, and it turns your HTS data into a systems biology story.

Computational breath analysis

MIMA is a method for the alignment of GC/MS data and MCC/IMS data facilitating the identification of biomarker molecules.

 

IMSDB is database system and machine learning user interface for biomarker mining and identification in breath data from GC/MS and MCC/IMS data sets.

 

Carotta is a method for the identification of hidden confounding factors in breath data.

Drug repositioning

N-Clue is a tool for n-clustering of data of different types simultaneously. We applied it to what we call "gene-drug-disease triangulation" using gene-disease associations, drug targets, and medication databases. Afterwards, we coupled it to a text mining system to predict ca. 10K novel tricks for existing drugs, of which we find ca. 10% (1K) to be mentioned in pubmed abstracts in a "treatment context".

Large-scale clustering of big biomedical data sets

Transitivity Clustering is a tool for solving the Weighted Graph Cluster Editing Problem using a force-based heuristic. It comes with several extensions. TransClustMV, for instance, works with missing values.

 

ClustEval is a platform for the evaluation of clustering tools in many different biomedical settings.

Bi-clustering and bi-cluster editing

BiCluE is a software for weighted bi-cluster editing. It can be applied to identify groups of phenotypes that are associated with sets of genotypes (sampled from e.g. GWAS data).

 

BiForce is an extension of BiCLuE. It was applied to bi-cluster gene expression data sets to find sets of genes correlating with sets of conditions.

Network alignment

GEDEVO is a method for pairwise network (graph) alignment using optimizing a global criterion: minimal graph edit distance, i.e. the minimal number of node/edge additions/deletions to transform one graph into another one. GEDEVO-M is an extension of GEDEVO for the alignment of multiple (>2) input networks. And NABEECO solves network alignment using Bee Colony Optimization.

 

CytoGEDEVO is a Cytoscape implementation of the pairwise network alignment algorithm of GEDEVO, which allows the direct integration of external node similarities (e.g. BLAST or graphlets).

Gene regulatory networks

CoryneRegNet is the international reference database for corynebacterial transcriptional gene regulatory interactions.

EhecRegNet is a database of gene regulations conserved between E. coli K12 and human pathogenic EHEC strains.

PetriScape is a tool for discrete Petri net simulation of biological networks in Cytoscape.

Bacterial genomic islands

LiSSI is a java tool for predicting life-style-specific genomic islands through evolutionary conservation. It is divided into three sequentially executed modules: evolutionary sequence analysis, island detection, and machine learning.

 

PIPS is a tool specifically for the prediction of pathogenicity islands using sequence features.

GIPSy extends on the PIPS idea and implements a software for using genomic features to predict pathogenicity islands as well metabolic and resistance islands in bacterial genomes.

Reverse phase protein array (RPPA) analysis

MIRACLE is an online platform for microarray R-based analysis of complex lysate experiments. It is bridging the gap between spotting and array analysis by conveniently keeping track of sample information. Data processing includes correction of staining bias, estimation of protein concentration from response curves, normalization for total protein amount per sample and statistical evaluation.

Laboratory managment and sample tracking

OpenLabFramework with its extension OpenLabNotes is an open-source laboratory information management system (LIMS) intended for advanced sample management in small to mid-sized laboratories. It has been developed with focus on the management of vector clone and genetically engineered cell lines.

Please reload

 

Chair of Experimental Bioinformatics

TUM School of Life Sciences Weihenstephan

Technical University of Munich

Tel: +49-8161-71-2136

E-Mail: exbio[a.t_))wzw.tum.de

Office: OG-L06