MOTIVATION: Protein-protein interaction datasets, which can be modeled as networks, constitute an essential layer in multi-omics approach to biomedical knowledge. This representation gives insight into molecular pathways, help to uncover novel potential drug targets or predict a therapy outcome. Nevertheless, the data that constitute such systems are frequently incomplete, error-prone and biased by scientific trends. Implementation of methods for detection of such shortcomings could improve protein-protein interaction data analysis. RESULTS: We performed topological analysis of three protein-protein interaction networks (PPINs) from IntAct Molecular Database, regarding cancer, Parkinson's disease (two most common subjects in PPINs analysis) and Human Reference Interactome. The data collections were shown to be often biased by scientific interests, which highly impact the networks structure. This may obscure correct systematic biological interpretation of the protein-protein interactions and limit their application potential. As a solution to this problem, we propose a set of topological methods for the bias detection, which performed in the first step provides more objective biological conclusions regarding protein-protein interactions and their multi-omics consequences. AVAILABILITY AND IMPLEMENTATION: A user-friendly tool Extensive Tool for Network Analysis (ETNA) is available on https://github.com/AlicjaNowakowska/ETNA. The software includes a graphical Colab notebook: https://githubtocolab.com/AlicjaNowakowska/ETNA/blob/main/ETNAColab.ipynb. CONTACT: [email protected] or [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Motivation: Residue-residue contact prediction through direct coupling analysis has reached impressive accuracy, but yet higher accuracy will be needed to allow for routine modelling of protein structures. One way to improve the prediction accuracy is to filter predicted contacts using knowledge about the particular protein of interest or knowledge about protein structures in general. Results: We focus on the latter and discuss a set of filters that can be used to remove false positive contact predictions. Each filter depends on one or a few cut-off parameters for which the filter performance was investigated. Combining all filters while using default parameters resulted for a test set of 851 protein domains in the removal of 29% of the predictions of which 92% were indeed false positives. Availability and implementation: All data and scripts are available at http://comprec-lin.iiar.pwr.edu.pl/FPfilter/. Supplementary information: Supplementary data are available at Bioinformatics online.

MOTIVATION: Apart from meta-predictors, most of today's methods for residue-residue contact prediction are based entirely on Direct Coupling Analysis (DCA) of correlated mutations in multiple sequence alignments (MSAs). These methods are on average ∼40% correct for the 100 strongest predicted contacts in each protein. The end-user who works on a single protein of interest will not know if predictions are either much more or much less correct than 40%, which is especially a problem if contacts are predicted to steer experimental research on that protein. RESULTS: We designed a regression model that forecasts the accuracy of residue-residue contact prediction for individual proteins with an average error of 7 percentage points. Contacts were predicted with two DCA methods (gplmDCA and PSICOV). The models were built on parameters that describe the MSA, the predicted secondary structure, the predicted solvent accessibility and the contact prediction scores for the target protein. Results show that our models can be also applied to the meta-methods, which was tested on RaptorX. AVAILABILITY AND IMPLEMENTATION: All data and scripts are available from http://comprec-lin.iiar.pwr.edu.pl/dcaQ/. CONTACT: [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

MOTIVATION: The recently developed direct coupling analysis (DCA) method has greatly improved the accuracy with which residue-residue contacts can be predicted from multiple sequence alignments. Contact prediction accuracy, though, is still often not sufficient for complete ab initio protein structure prediction. DCA can, however, support protein structure studies in several ways. RESULTS: We show that DCA can select the better structure from among properly folded and misfolded variants. This idea was tested by comparing obsolete PDB files with their more correctly folded successors and by the comparison of structures with deliberately misfolded decoy models from the Decoys 'R' Us database. The DCA method systematically predicts more contacts for properly folded structures than for misfolded ones. The method works much better for X-ray structures than for NMR structures. AVAILABILITY AND IMPLEMENTATION: All data are available from http://comprec-lin.iiar.pwr.edu.pl/dcaVSmisfolds/ and http://swift.cmbi.ru.nl/dcaVSmisfolds/ . CONTACT: [email protected] . SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

UNLABELLED: Analyses of amyloidogenic sequence fragments are essential in studies of neurodegenerative diseases. However, there is no one internet dataset that collects all the sequences that have been investigated for their amyloidogenicity. Therefore, we have created the AmyLoad website which collects the amyloidogenic sequences from all major sources. The website allows for filtration of the fragments and provides detailed information about each of them. Registered users can both personalize their work with the website and submit their own sequences into the database. To maintain database reliability, submitted sequences are reviewed before making them available to the public. Finally, we re-implemented several amyloidogenic sequence predictors, thus the AmyLoad website can be used as a sequence analysis tool. We encourage researchers working on amyloid proteins to contribute to our service. AVAILABILITY AND IMPLEMENTATION: The AmyLoad website is freely available at http://comprec-lin.iiar.pwr.edu.pl/amyload/. CONTACT: [email protected].

BACKGROUND: Amyloids are proteins capable of forming fibrils whose intramolecular contact sites assume densely packed zipper pattern. Their oligomers can underlie serious diseases, e.g. Alzheimer's and Parkinson's diseases. Recent studies show that short segments of aminoacids can be responsible for amyloidogenic properties of a protein. A few hundreds of such peptides have been experimentally found but experimental testing of all candidates is currently not feasible. Here we propose an original machine learning method for classification of aminoacid sequences, based on discovering a segment with a discriminative pattern of site-specific co-occurrences between sequence elements. The pattern is based on the positions of residues with correlated occurrence over a sliding window of a specified length. The algorithm first recognizes the most relevant training segment in each positive training instance. Then the classification is based on maximal distances between co-occurrence matrix of the relevant segments in positive training sequences and the matrix from negative training segments. The method was applied for studying sequences of aminoacids with regard to their amyloidogenic properties. RESULTS: Our method was first trained on available datasets of hexapeptides with the amyloidogenic classification, using 5 or 6-residue sliding windows. Depending on the choice of training and testing datasets, the area under ROC curve obtained the value up to 0.80 for experimental, and 0.95 for computationally generated (with 3D profile method) datasets. Importantly, the results on 5-residue segments were not significantly worse, although the classification required that algorithm first recognized the most relevant training segments. The dataset of long sequences, such as sup35 prion and a few other amyloid proteins, were applied to test the method and gave encouraging results. Our web tool FISH Amyloid was trained on all available experimental data 4-10 residues long, offers prediction of amyloidogenic segments in protein sequences. CONCLUSIONS: We proposed a new original classification method which recognizes co-occurrence patterns in sequences. The method reveals characteristic classification pattern of the data and finds the segments where its scoring is the strongest, also in long training sequences. Applied to the problem of amyloidogenic segments recognition, it showed a good potential for classification problems in bioinformatics.

BACKGROUND: Amyloids are proteins capable of forming aberrant intramolecular contact sites, characteristic of beta zipper configuration. Amyloids can underlie serious health conditions, e.g. Alzheimer's or Parkinson's diseases. It has been proposed that short segments of amino acids can be responsible for protein amyloidogenicity, but no more than two hundred such hexapeptides have been experimentally found. The authors of the computational tool Pafig published in BMC Bioinformatics a method for extending the amyloid hexapeptide dataset that could be used for training and testing models. They assumed that all hexapeptides belonging to an amyloid protein can be regarded as amylopositive, while those from proteins never reported as amyloid are always amylonegative. Here we show why the above described method of extending datasets is wrong and discuss the reasons why the incorrect data could lead to falsely correct classification. RESULTS: The amyloid classification of hexapeptides by Pafig was confronted with the classification results from different state of the art computational methods and the outputs of all methods were studied by clustering analysis. The clustering methods show that Pafig is an outlier with regard to other approaches. Our study of the statistical patterns of its training and testing datasets showed a strong bias towards STVIIE hexapeptide in their positive part. Different statistical patterns of seemingly amylo-positive and -negative hexapeptides allow for a repeatable classification, which is not related to amyloid propensity of the hexapetides. CONCLUSIONS: Our study on recognition of amyloid hexapeptides showed that occurrence of incidental patterns in wrongly selected datasets can produce falsely correct results of classification. The assumption that all hexapeptides belonging to amyloid protein can be regarded as amylopositive and those from proteins never reported as amyloid are always amylonegative is not supported by any other computational method. This is in line with experimental observations that amyloid propensity of a full protein can result from only one amyloidogenic fragment in this protein, while the occurrence of amyliodogenic part that is well hidden inside the protein may never lead to fibril formation. This leads to the conclusion that Pafig does not provide correct classification with regard to amyloidogenicity.

BACKGROUND: Amyloids are proteins capable of forming fibrils. Many of them underlie serious diseases, like Alzheimer disease. The number of amyloid-associated diseases is constantly increasing. Recent studies indicate that amyloidogenic properties can be associated with short segments of aminoacids, which transform the structure when exposed. A few hundreds of such peptides have been experimentally found. Experimental testing of all possible aminoacid combinations is currently not feasible. Instead, they can be predicted by computational methods. 3D profile is a physicochemical-based method that has generated the most numerous dataset - ZipperDB. However, it is computationally very demanding. Here, we show that dataset generation can be accelerated. Two methods to increase the classification efficiency of amyloidogenic candidates are presented and tested: simplified 3D profile generation and machine learning methods. RESULTS: We generated a new dataset of hexapeptides, using more economical 3D profile algorithm, which showed very good classification overlap with ZipperDB (93.5%). The new part of our dataset contains 1779 segments, with 204 classified as amyloidogenic. The dataset of 6-residue sequences with their binary classification, based on the energy of the segment, was applied for training machine learning methods. A separate set of sequences from ZipperDB was used as a test set. The most effective methods were Alternating Decision Tree and Multilayer Perceptron. Both methods obtained area under ROC curve of 0.96, accuracy 91%, true positive rate ca. 78%, and true negative rate 95%. A few other machine learning methods also achieved a good performance. The computational time was reduced from 18-20 CPU-hours (full 3D profile) to 0.5 CPU-hours (simplified 3D profile) to seconds (machine learning). CONCLUSIONS: We showed that the simplified profile generation method does not introduce an error with regard to the original method, while increasing the computational efficiency. Our new dataset proved representative enough to use simple statistical methods for testing the amylogenicity based only on six letter sequences. Statistical machine learning methods such as Alternating Decision Tree and Multilayer Perceptron can replace the energy based classifier, with advantage of very significantly reduced computational time and simplicity to perform the analysis. Additionally, a decision tree provides a set of very easily interpretable rules.

BACKGROUND: Experimental determination of protein 3D structures is expensive, time consuming and sometimes impossible. A gap between number of protein structures deposited in the World Wide Protein Data Bank and the number of sequenced proteins constantly broadens. Computational modeling is deemed to be one of the ways to deal with the problem. Although protein 3D structure prediction is a difficult task, many tools are available. These tools can model it from a sequence or partial structural information, e.g. contact maps. Consequently, biologists have the ability to generate automatically a putative 3D structure model of any protein. However, the main issue becomes evaluation of the model quality, which is one of the most important challenges of structural biology. RESULTS: GOBA--Gene Ontology-Based Assessment is a novel Protein Model Quality Assessment Program. It estimates the compatibility between a model-structure and its expected function. GOBA is based on the assumption that a high quality model is expected to be structurally similar to proteins functionally similar to the prediction target. Whereas DALI is used to measure structure similarity, protein functional similarity is quantified using standardized and hierarchical description of proteins provided by Gene Ontology combined with Wang's algorithm for calculating semantic similarity. Two approaches are proposed to express the quality of protein model-structures. One is a single model quality assessment method, the other is its modification, which provides a relative measure of model quality. Exhaustive evaluation is performed on data sets of model-structures submitted to the CASP8 and CASP9 contests. CONCLUSIONS: The validation shows that the method is able to discriminate between good and bad model-structures. The best of tested GOBA scores achieved 0.74 and 0.8 as a mean Pearson correlation to the observed quality of models in our CASP8 and CASP9-based validation sets. GOBA also obtained the best result for two targets of CASP8, and one of CASP9, compared to the contest participants. Consequently, GOBA offers a novel single model quality assessment program that addresses the practical needs of biologists. In conjunction with other Model Quality Assessment Programs (MQAPs), it would prove useful for the evaluation of single protein models.