BACKGROUND: Learning the representation of molecules is crucial for drug discovery but is often hindered by the scarcity of labeled experimental data, which limits the performance of supervised machine learning models. While self-supervised learning (SSL) offers a solution by leveraging vast unlabeled chemical databases, many existing methods focus on learning from either local structural information or global molecular properties, but not both simultaneously. We introduce MoCL-GAT, a novel contrastive and transfer learning-based SSL framework that addresses this gap by simultaneously learning from two complementary objectives. It combines a local contrastive task on molecular subgraphs to capture fine-grained chemical environments with a global predictive task to learn holistic molecular descriptors. This dual-objective approach, powered by a Graph Attention Network, is designed to create more robust, versatile, and transferable molecular representations. RESULTS: Pre-trained on 1.9 million compounds, MoCL-GAT was fine-tuned on diverse benchmarks. It achieved state-of-the-art performance on molecular property prediction tasks, with an AUROC of 0.928 on BBBP and 0.768 on SIDER, and top-ranking RMSEs of 0.570 for ESOL and 1.818 for FreeSolv. Critically, fine-tuned models consistently and significantly outperformed models trained from scratch, confirming the value of pre-training. CONCLUSIONS: These results validate that MoCL-GAT's dual-objective approach learns highly effective and transferable representations, enabling more accurate and data-efficient predictions for key cheminformatics challenges. The source code for MoCL-GAT is publicly available on Zenodo at https://doi.org/10.5281/zenodo.16927285 .

As the number of protein sequences increases in biological databases, computational methods are required to provide accurate functional annotation with high coverage. Although several machine learning methods have been proposed for this purpose, there are still two main issues: (i) construction of reliable positive and negative training and validation datasets, and (ii) fair evaluation of their performances based on predefined experimental settings. To address these issues, we have developed ProFAB: Open Protein Functional Annotation Benchmark, which is a platform providing an infrastructure for a fair comparison of protein function prediction methods. ProFAB provides filtered and preprocessed protein annotation datasets and enables the training and evaluation of function prediction methods via several options. We believe that ProFAB will be useful for both computational and experimental researchers by enabling the utilization of ready-to-use datasets and machine learning algorithms for protein function prediction based on Gene Ontology terms and Enzyme Commission numbers. ProFAB is available at https://github.com/kansil/ProFAB and https://profab.kansil.org.

MOTIVATION: Utilizing AI-driven approaches for drug-target interaction (DTI) prediction require large volumes of training data which are not available for the majority of target proteins. In this study, we investigate the use of deep transfer learning for the prediction of interactions between drug candidate compounds and understudied target proteins with scarce training data. The idea here is to first train a deep neural network classifier with a generalized source training dataset of large size and then to reuse this pre-trained neural network as an initial configuration for re-training/fine-tuning purposes with a small-sized specialized target training dataset. To explore this idea, we selected six protein families that have critical importance in biomedicine: kinases, G-protein-coupled receptors (GPCRs), ion channels, nuclear receptors, proteases, and transporters. In two independent experiments, the protein families of transporters and nuclear receptors were individually set as the target datasets, while the remaining five families were used as the source datasets. Several size-based target family training datasets were formed in a controlled manner to assess the benefit provided by the transfer learning approach. RESULTS: Here, we present a systematic evaluation of our approach by pre-training a feed-forward neural network with source training datasets and applying different modes of transfer learning from the pre-trained source network to a target dataset. The performance of deep transfer learning is evaluated and compared with that of training the same deep neural network from scratch. We found that when the training dataset contains fewer than 100 compounds, transfer learning outperforms the conventional strategy of training the system from scratch, suggesting that transfer learning is advantageous for predicting binders to under-studied targets. AVAILABILITY AND IMPLEMENTATION: The source code and datasets are available at https://github.com/cansyl/TransferLearning4DTI. Our web-based service containing the ready-to-use pre-trained models is accessible at https://tl4dti.kansil.org.

Statistical and machine learning techniques based on relative abundances have been used to predict health conditions and to identify microbial biomarkers. However, high dimensionality, sparsity and the compositional nature of microbiome data represent statistical challenges. On the other hand, the taxon grouping allows summarizing microbiome abundance with a coarser resolution in a lower dimension, but it presents new challenges when correlating taxa with a disease. In this work, we present a novel approach that groups Operational Taxonomical Units (OTUs) based only on relative abundances as an alternative to taxon grouping. The proposed procedure acknowledges the compositional data making use of principal balances. The identified groups are called Principal Microbial Groups (PMGs). The procedure reduces the need for user-defined aggregation of $\textrm{OTU}$s and offers the possibility of working with coarse group of $\textrm{OTU}$s, which are not present in a phylogenetic tree. PMGs can be used for two different goals: (1) as a dimensionality reduction method for compositional data, (2) as an aggregation procedure that provides an alternative to taxon grouping for construction of microbial balances afterward used for disease prediction. We illustrate the procedure with a cirrhosis study data. PMGs provide a coherent data analysis for the search of biomarkers in human microbiota. The source code and demo data for PMGs are available at: https://github.com/asliboyraz/PMGs.

Bioinformatics (2010) doi: 10.1093/bioinformatics/btaa496 Co-author Tunca Doğan’s name was initially misspelled as Tunca Doğana. This has been corrected online.

SUMMARY: iBioProVis is an interactive tool for visual analysis of the compound bioactivity space in the context of target proteins, drugs and drug candidate compounds. iBioProVis tool takes target protein identifiers and, optionally, compound SMILES as input, and uses the state-of-the-art non-linear dimensionality reduction method t-Distributed Stochastic Neighbor Embedding (t-SNE) to plot the distribution of compounds embedded in a 2D map, based on the similarity of structural properties of compounds and in the context of compounds' cognate targets. Similar compounds, which are embedded to proximate points on the 2D map, may bind the same or similar target proteins. Thus, iBioProVis can be used to easily observe the structural distribution of one or two target proteins' known ligands on the 2D compound space, and to infer new binders to the same protein, or to infer new potential target(s) for a compound of interest, based on this distribution. Principal component analysis (PCA) projection of the input compounds is also provided, Hence the user can interactively observe the same compound or a group of selected compounds which is projected by both PCA and embedded by t-SNE. iBioProVis also provides detailed information about drugs and drug candidate compounds through cross-references to widely used and well-known databases, in the form of linked table views. Two use-case studies were demonstrated, one being on angiotensin-converting enzyme 2 (ACE2) protein which is Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Spike protein receptor. ACE2 binding compounds and seven antiviral drugs were closely embedded in which two of them have been under clinical trial for Coronavirus disease 19 (COVID-19). AVAILABILITY AND IMPLEMENTATION: iBioProVis and its carefully filtered dataset are available at https://ibpv.kansil.org/ for public use. CONTACT: [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

We motivate and describe the application of Hierarchical Dirichlet Process (HDP) models to the "soft" biclustering of gene expression data, in which we obtain modules (biclusters) where the affiliation of genes and samples with the modules are weighted, instead of being hard memberships. As a distinct contribution, we propose a method which HDP is informed with prior beliefs, significantly increasing the quality of the biclustering in terms of both the correctness of the number of modules inferred, and the precision of these modules, especially when evidence is sparse. We outline two such informed priors; one based on co-expression relationships inherent in the data, the other based on an externally provided regulatory network. We validate these results and compare the performance of our approach to Weighted Gene Correlation Network Analysis (WGCNA), another model that features weighted modules. We have, to this end, performed experiments on semi-synthetic data. The results show that HDP, with the addition of a well-informed prior, is able to capture the correct number of modules with increased accuracy. Furthermore, the model becomes robust to changes in the strength of the prior. We conclude by discussing these results and the benefits provided by our approach for gene expression analysis and network validation.

We describe a method of inferring join plans for a set of relation instances, in the absence of any metadata, such as attribute domains, attribute names, or constraints (e.g., keys or foreign keys). Our method enumerates the possible join plans in order of likelihood, based on the compatibility of a pair of columns and their suitability as join attributes (i.e. their appropriateness as keys). We outline two variants of the approach. The first variant is accurate but potentially time-consuming, especially for large relations that do not fit in memory. The second variant is an approximation of the former and hence less accurate, but is considerably more efficient, allowing the method to be used online, even for large relations. We provide experimental results showing how both forms scale in terms of performance as the number of candidate join attributes and the size of the relations increase. We also characterize the accuracy of the approximate variant with respect to the exact variant.

A sequence of queries submitted by a database user within a short period of time may have a single, illuminating explanation. In this paper we consider sequences of single-record queries, and attempt to guess what information their authors may be trying to accumulate. Query sequences may reflect clandestine intentions, where users attempt to avoid direct queries which may disclose their true interests, preferring instead to obtain the same information by means of sequences of smaller, less conspicuous, queries. Sequences of queries may also reflect attempts to circumvent retrieval restrictions , where users attempt to approximate information which is inaccessible, with sequences of legitimate requests (in the latter case, our explanations may lead database owners to either tighten access, or, conversely, to reorganize their interfaces to facilitate access). Because the true objective of a sequence may be clouded by the retrieval of spurious records, our approach considers all the possible aggregates that a user may accumulate with a sequence, and to rank them, search-engine style, according to their plausibility as retrieval objectives. Our method is probabilistic in nature and postulates that the likelihood that a set of records is the true objective of the user is inverse proportional to the likelihood that this set results from random selection. Our method is shown to have good performance even in the presence of noise (spurious records) as high as 40–50%.