Biology subjects

Ferdinandy, P.

Publications and source records attributed to Ferdinandy, P..

3 recordsLinked to original sources

bioRxiv2025

Infarctsize-AI: an efficient infarct size image analysis tool for small rodent myocardial infarction studies

BackgroundMyocardial infarct size (IS) is the gold standard end-point in shorth-term studies on cardioprotection. However, IS quantification in rodent models with standard Evans Blue and 2,3,5-triphenyltetrazolium chloride (TTC) staining is time-consuming and prone to inter-observer variance. Therefore, we aimed to develop an artificial intelligence (AI)-based application to reduce time and inter-observer variability of IS analysis in rodent acute myocardial infarction (MI) models. MethodsWe used TTC/Evans blue-stained heart slice images of independent laboratories from previously published projects. Rat (n = 325 and 248 slices) and mouse (n = 77 slices) datasets were used to train deep learning segmentation models with three different neural network architectures, which were combined into a single AI analysis. AI analysis was compared with manual analysis on rat data from a training laboratory (internal data, n = 496 slices, n = 41 whole-hearts) and data from independent laboratories (external data, n = 60 and 62 slices). Additionally, two independent evaluators performed manual and AI-assisted analysis, consisting of AI-analysis and its manual correction, on internal (n = 36 slices) and external data (n = 37 slices). ResultsLins concordance correlation coefficient (CCC) between IS/AAR values from manual and AI analysis was 0.844 with 95% CI of [0.814; 0.869] for images of internal data heart slices. On external data heart slices, AI accurately annotated slice area and AAR but failed to annotate infarcted area. On internal whole-heart data, CCC between AI and AI-assisted IS/AAR was 0.894 with 95% CI of [0.812; 0.942]. AI-assisted analysis reduced evaluation time on both internal and external datasets and increased region overlap for AAR between the two independent evaluators on dependent data. ConclusionsAI-assisted analysis significantly reduced analysis time and inter-observer variability. For optimal performance, lab-specific AI training is recommended. Infarctsize-AI is available at https://infarctsize.com. Translational perspectiveMyocardial infarct size (IS) is the gold-standard end-point in shorth-term studies to assess potential cardioprotective therapies against acute myocardial infarction (AMI). However, IS quantification in rodent AMI models is time-consuming and prone to inter-observer variance. Therefore, we developed an AI-based software that can reduce analysis time and inter-observer variability and facilitate documentation, which facilitates the clinical translation of potential cardioprotective therapies. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=34 SRC="FIGDIR/small/688527v1_ufig1.gif" ALT="Figure 1"> View larger version (15K): org.highwire.dtl.DTLVardef@187c938org.highwire.dtl.DTLVardef@1ab8bb4org.highwire.dtl.DTLVardef@160dc68org.highwire.dtl.DTLVardef@264617_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2025

Contrastive learning of adverse events to provide effective and interpretable vector representations for machine-assisted pharmacovigilance

Post-marketing surveillance is crucial for drug safety, yet the tools of pharmacovigilance rely solely on text-based data that may limit the applicability of contemporary machine learning methodologies in the support of decision making. Here, we adapt contrastive learning algorithms to generate adverse event vector representations from spontaneous reports to serve as general machine-readable resources for pharmacovigilance applications. We present comprehensive analyses of the resulting representations through density-based clustering, semantic evaluation and comparison of multivariate dispersions, revealing patterns that reflect both functional and causal relations of the adverse events while also capturing drug-safety related information better than existing medical terminologies and encoder-only large language models (LLMs). Furthermore, we demonstrate the applicability of the representations as input features in our downstream model, outperforming the reporting odds ratio method commonly used by regulatory agencies (AUROC: 0.88 vs 0.75) and LLM-based representations (AUROC: 0.88 vs 0.83) on drug-event causality prediction benchmarks. As such, this is the first demonstration of an interpretable adverse event vector representation that can be utilized for training arbitrary models, enabling wider and more effective applications of machine learning in pharmacovigilance.

pharmacology and toxicology↗

bioRxiv2021

Assessment of community efforts to advance computational prediction of protein-protein interactions

Comprehensive insights from the human protein-protein interaction (PPI) network, known as the human interactome, can provide important insights into the molecular mechanisms of complex biological processes and diseases. Despite the remarkable experimental efforts undertaken to date to determine the structure of the human interactome, many PPIs remain unmapped. Computational approaches, especially network-based methods, can facilitate the identification of new PPIs. Many such approaches have been proposed. However, a systematic evaluation of existing network-based methods in predicting PPIs is still lacking. Here, we report community efforts initiated by the International Network Medicine Consortium to benchmark the ability of 24 representative network-based methods to predict PPIs across five different interactomes, including a synthetic interactome generated by the duplication-mutation-complementation model, and the interactomes of four different organisms: A. thaliana, C. elegans, S. cerevisiae, and H. sapiens. We selected the top-seven methods through a computational validation on the human interactome. We next experimentally validated their top-500 predicted PPIs (in total 3,276 predicted PPIs) using the yeast two-hybrid assay, finding 1,177 new human PPIs (involving 633 proteins). Our results indicate that task-tailored similarity-based methods, which leverage the underlying network characteristics of PPIs, show superior performance over other general link prediction methods. Through experimental validation, we confirmed that the top-ranking methods show promising performance externally. For example, from the top 500 PPIs predicted by an advanced similarity-base method [MPS(B&T)], 430 were successfully tested by Y2H with 376 testing positive, yielding a precision of 87.4%. These results establish advanced similarity-based methods as powerful tools for the prediction of human PPIs.

systems biology↗