Comparison of six breast cancer classifiers using qPCR.

MOTIVATION: Several gene expression-based risk scores and subtype classifiers for breast cancer were developed to distinguish high- and low-risk patients. Evaluating the performance of these classifiers helps to decide which classifiers should be used in clinical practice for personal therapeutic recommendations. So far, studies that compared multiple classifiers in large independent patient cohorts mostly used microarray measurements. qPCR-based classifiers were not included in the comparison or had to be adapted to the different experimental platforms. RESULTS: We used a prospective study of 726 early breast cancer patients from seven certified German breast cancer centers. Patients were treated according to national guidelines and the expressions of 94 selected genes were measured by the mid-throughput qPCR platform Fluidigm. Clinical and pathological data including outcome over five years is available. Using these data, we could compare the performance of six classifiers (scmgene and research versions of PAM50, ROR-S, recurrence score, EndoPredict and GGI). Similar to other studies, we found a similar or even higher concordance between most of the classifiers and most were also able to differentiate high- and low-risk patients. The classifiers that were originally developed for microarray data still performed similarly using the Fluidigm data. Therefore, Fluidigm can be used to measure the gene expressions needed by several classifiers for a large cohort with little effort. In addition, we provide an interactive report of the results, which enables a transparent, in-depth comparison of classifiers and their prediction of individual patients. AVAILABILITY AND IMPLEMENTATION: https://services.bio.ifi.lmu.de/pia/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Multi-Reference Spectral Library Yields Almost Complete Coverage of Heterogeneous LC-MS/MS Data Sets.

Spectral libraries play a central role in the analysis of data-independent-acquisition (DIA) proteomics experiments. A main assumption in current spectral library tools is that a single characteristic intensity pattern (CIP) suffices to describe the fragmentation of a peptide in a particular charge state (peptide charge pair). However, we find that this is often not the case. We carry out a systematic evaluation of spectral variability over public repositories and in-house data sets. We show that spectral variability is widespread and partly occurs under fixed experimental conditions. Using clustering of preprocessed spectra, we derive a limited number of multiple characteristic intensity patterns (MCIPs) for each peptide charge pair, which allow almost complete coverage of our heterogeneous data set without affecting the false discovery rate. We show that a MCIP library derived from public repositories performs in most cases similar to a "custom-made" spectral library, which has been acquired under identical experimental conditions as the query spectra. We apply the MCIP approach to a DIA data set and observe a significant increase in peptide recognition. We propose the MCIP approach as an easy-to-implement addition to current spectral library search engines and as a new way to utilize the data stored in spectral repositories.

RelExplain-integrating data and networks to explain biological processes.

MOTIVATION: The goal of many genome-wide experiments is to explain the changes between the analyzed conditions. Typically, the analysis is started with a set of differential genes DG and the first step is to identify the set of relevant biological processes BP . Current enrichment methods identify the involved biological process via statistically significant overrepresentation of differential genes in predefined sets, but do not further explain how the differential genes interact with each other or which other genes might be important for the enriched process. Other network-based methods determine subnetworks of interacting genes containing many differential genes, but do not employ process knowledge for a more focused analysis. RESULTS: RelExplain is a method to analyze a given biological process bp (e.g. identified by enrichment) in more detail by computing an explanation using the measured DG and a given network. An explanation is a subnetwork that contains the differential genes in the process bp and connects them in the best way given the experimental data using also genes that are not differential or not in bp . RelExplain takes into account the functional annotations of nodes and the edge consistency of the measurements. Explanations are compact networks of the relevant part of the bp and additional nodes that might be important for the bp . Our evaluation showed that RelExplain is better suited to retrieve manually curated subnetworks from unspecific networks than other algorithms. The interactive RelExplain tool allows to compute and inspect sub-optimal and alternative optimal explanations. AVAILABILITY AND IMPLEMENTATION: A webserver is available at https://services.bio.ifi.lmu.de/relexplain . CONTACT: berchtold@bio.ifi.lmu.de. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Deletion of the transcription factors Hsf1, Msn2 and Msn4 in yeast uncovers transcriptional reprogramming in response to proteotoxic stress.

The response to proteotoxic stresses such as heat shock allows organisms to maintain protein homeostasis under changing environmental conditions. We asked what happens if an organism can no longer react to cytosolic proteotoxic stress. To test this, we deleted or depleted, either individually or in combination, the stress-responsive transcription factors Msn2, Msn4, and Hsf1 in Saccharomyces cerevisiae. Our study reveals a combination of survival strategies, which together protect essential proteins. Msn2 and 4 broadly reprogram transcription, triggering the response to oxidative stress, as well as biosynthesis of the protective sugar trehalose and glycolytic enzymes, while Hsf1 mainly induces the synthesis of molecular chaperones and reverses the transcriptional response upon prolonged mild heat stress (adaptation).

Detection and correction of probe-level artefacts on microarrays.

BACKGROUND: A recent large-scale analysis of Gene Expression Omnibus (GEO) data found frequent evidence for spatial defects in a substantial fraction of Affymetrix microarrays in the GEO. Nevertheless, in contrast to quality assessment, artefact detection is not widely used in standard gene expression analysis pipelines. Furthermore, although approaches have been proposed to detect diverse types of spatial noise on arrays, the correction of these artefacts is mostly left to either summarization methods or the corresponding arrays are completely discarded. RESULTS: We show that state-of-the-art robust summarization procedures are vulnerable to artefacts on arrays and cannot appropriately correct for these. To address this problem, we present a simple approach to detect artefacts with high recall and precision, which we further improve by taking into account the spatial layout of arrays. Finally, we propose two correction methods for these artefacts that either substitute values of defective probes using probeset information or filter corrupted probes. We show that our approach can identify and correct defective probe measurements appropriately and outperforms existing tools. CONCLUSIONS: While summarization is insufficient to correct for defective probes, this problem can be addressed in a straightforward way by the methods we present for identification and correction of defective probes. As these methods output CEL files with corrected probe values that serve as input to standard normalization and summarization procedures, they can be easily integrated into existing microarray analysis pipelines as an additional pre-processing step. An R package is freely available from http://www.bio.ifi.lmu.de/artefact-correction.

Coevolution of ABC transporters and two-component regulatory systems as resistance modules against antimicrobial peptides in Firmicutes Bacteria.

In Firmicutes bacteria, ATP-binding cassette (ABC) transporters have been recognized as important resistance determinants against antimicrobial peptides. Together with neighboring two-component systems (TCSs), which regulate their expression, they form specific detoxification modules. Both the transport permease and sensor kinase components show unusual domain architecture: the permeases contain a large extracellular domain, while the sensor kinases lack an obvious input domain. One of the best-characterized examples is the bacitracin resistance module BceRS-BceAB of Bacillus subtilis. Strikingly, in this system, the ABC transporter and TCS have an absolute mutual requirement for each other in both sensing of and resistance to bacitracin, suggesting a novel mode of signal transduction in which the transporter constitutes the actual sensor. We identified over 250 such BceAB-like ABC transporters in the current databases. They occurred almost exclusively in Firmicutes bacteria, and 80% of the transporters were associated with a BceRS-like TCS. Phylogenetic analyses of the permease and sensor kinase components revealed a tight evolutionary correlation. Our findings suggest a direct regulatory interaction between the ABC transporters and TCSs, mediating communication between both components. Based on their observed coclustering and conservation of response regulator binding sites, we could identify putative corresponding two-component systems for transporters lacking a regulatory system in their immediate neighborhood. Taken together, our results show that these types of ABC transporters and TCSs have coevolved to form self-sufficient detoxification modules against antimicrobial peptides, widely distributed among Firmicutes bacteria.

YESdb: integrative analysis of environmental stress in yeast.

The stress response in the model organisms Saccharomyces cerevisiae is a well-studied system for which many data sets are available. Already in 2000, it was discovered that yeast cells trigger a similar transcriptional response when different types of stress are applied. However, the exact regulatory mechanisms and differences between the different types of stress are still not understood. Here, we present the Yeast Environmental Stress database (YESdb), a database containing all high-throughput experiments measuring various kinds of stress in yeast. The goal of the database is to allow the user to execute complex, integrative analyses of selected data sets, e.g. the comparison of measurements of the same stress using different platforms or differences between strains, stress strengths or types of stress. The analyses can be visualized in various ways and can be compiled into interactive reports to summarize and communicate the results. The data sets are available as differential conditions (typically stressed vs control), which are grouped to time or concentration series when multiple measurements over time or concentrations are done in one experiment. An annotation ontology has been constructed to annotate the data sets with the type, duration and strength of the applied stress, the used strain and experimental platform as well as the publication date. These annotations can easily be combined to select all relevant data sets for an analysis. YESdb allows to construct and execute Petri net-based workflows to perform predefined and custom analyses. E.g. to compare two types of stress (e.g. salt vs oxidative stress), the corresponding data sets are selected from the database, the consistently changed genes are defined and combined and the shared genes are characterized by enrichment analysis. A broad collection of visualizations is available most of which are also interactive. The results of all analyses can be summarized in an interactive report. Visualizations of individual steps (transitions) of YESdb workflows can be automatically added to this report or customized visualizations as well as interpretive text can manually be added to the report. Overall, YESdb aims at making all published data sets on yeast stress immediately available and comparable for integrated analysis of data sets and sets of genes in order to identify and assess hypotheses and mechanisms.

The Heat Shock Response in Yeast Maintains Protein Homeostasis by Chaperoning and Replenishing Proteins.

Life is resilient because living systems are able to respond to elevated temperatures with an ancient gene expression program called the heat shock response (HSR). In yeast, the transcription of hundreds of genes is upregulated at stress temperatures. Besides stress protection conferred by chaperones, the function of the majority of the upregulated genes under stress has remained enigmatic. We show that those genes are required to directly counterbalance increased protein turnover at stress temperatures and to maintain the metabolism. This anaplerotic reaction together with molecular chaperones allows yeast to efficiently buffer proteotoxic stress. When the capacity of this system is exhausted at extreme temperatures, aggregation processes stop translation and growth pauses. The emerging concept is that the HSR is modular with distinct programs dependent on the severity of the stress.

Evaluating Transcription Factor Activity Changes by Scoring Unexplained Target Genes in Expression Data.

Several methods predict activity changes of transcription factors (TFs) from a given regulatory network and measured expression data. But available gene regulatory networks are incomplete and contain many condition-dependent regulations that are not relevant for the specific expression measurement. It is not known which combination of active TFs is needed to cause a change in the expression of a target gene. A method to systematically evaluate the inferred activity changes is missing. We present such an evaluation strategy that indicates for how many target genes the observed expression changes can be explained by a given set of active TFs. To overcome the problem that the exact combination of active TFs needed to activate a gene is typically not known, we assume a gene to be explained if there exists any combination for which the predicted active TFs can possibly explain the observed change of the gene. We introduce the i-score (inconsistency score), which quantifies how many genes could not be explained by the set of activity changes of TFs. We observe that, even for these minimal requirements, published methods yield many unexplained target genes, i.e. large i-scores. This holds for all methods and all expression datasets we evaluated. We provide new optimization methods to calculate the best possible (minimal) i-score given the network and measured expression data. The evaluation of this optimized i-score on a large data compendium yields many unexplained target genes for almost every case. This indicates that currently available regulatory networks are still far from being complete. Both the presented Act-SAT and Act-A* methods produce optimal sets of TF activity changes, which can be used to investigate the difficult interplay of expression and network data. A web server and a command line tool to calculate our i-score and to find the active TFs associated with the minimal i-score is available from https://services.bio.ifi.lmu.de/i-score.