Biology subjects

Perera-Lluna, A.

Publications and source records attributed to Perera-Lluna, A..

6 recordsLinked to original sources

bioRxiv2026

singIST: an R/Bioconductor library and Quarto dashboard for automated single-cell comparative transcriptomics analysis ofdisease models and humans

Preclinical disease models often diverge from human pathophysiology at single-cell resolution, complicating model selection and limiting translational value. We present singIST, an R/Bioconductor package for quantitative and explainable comparison of disease model scRNA-seq data against a human reference. For each superpathway, singIST fits an adaptive sparse multi-block PLS-DA model on human pseudobulk expression, integrated one-to-one orthology and cell type mapping, and translates model fold changes into the human expression space to compute signed recapitulation at the superpathway, cell type, and gene levels. To streamline interpretation and reporting, we provide singIST Visualizer, a companion Quarto/Shiny dashboard that loads singIST outputs and offers interactive exploration with export ready plots and tables, avoiding manual figure coding across many superpathways and models. We demonstrate the workflow. We illustrate an end-to-end workflow on an oxazolone mouse model against a human atopic dermatitis reference for two representative pathways: Dendritic Cells in regulating Th1/Th2 Development [BIOCARTA] and Cytokine-cytokine receptor interaction [KEGG]. singIST is distributed under the MIT License via Bioconductor, and the Visualizer is available on GitHub.

bioinformatics↗

bioRxiv2024

singIST: an integrative method for comparative single-cell trancriptomics between disease models and humans

MotivationDisease models serve as fundamental tools in drug discovery and early-stage drug development. However, these models are not a perfect reflection of human disease, and selecting a suitable model can be challenging. Existing computational approaches for molecular validation of pathophysiological resemblance to human conditions at single-cell resolution remain limited. Although quantitative computational methods exist to inform this selection, they are very limited at the single-cell resolution, which can be critical for model selection. Quantifying the resemblance of disease models to the human condition with single-cell technologies in an explainable, integrative, and generalizable manner remains a significant challenge. ResultsWe present singIST, a computational method for comparative single-cell transcriptomics analysis between disease models and human conditions. singIST provides explainable quantitative measures on disease model similarity to human condition at both pathway and cell type levels, highlighting the importance of each gene in the latter. These measures account for orthology, cell type presence in the disease model, cell type and gene importance in human condition, and gene changes in the disease model measured as fold change. This is achieved within a unifying framework that controls for the intrinsic complexities of single-cell data. We tested our method using three well-characterized murine models of moderate-to-severe Atopic Dermatitis, demonstrating its ability to recapitulate established biological knowledge while generating novel hypothesis through pathway-level analysis. Availability and implementationR library at https://github.com/DataScienceRD-Almirall/singIST and implementation code at https://github.com/DataScienceRD-Almirall/singIST_paper_results Author SummaryIn this study, we set out a method to improve how researchers evaluate disease models - tools that are essential for understanding human illness and developing new treatments. A major limitation of current methods is that they overlook the fine-scale differences between disease models and human conditions at the single-cell level. Yet, many diseases are driven by changes that occur in specific cell types. To address this, we developed singIST, a method that allows us to compare disease models to human diseases using single-cell gene expression data. What makes singIST unique is its ability to provide clear, interpretable measures of similarity between a disease model and the human condition, specifically at the level of cell types and groups of functionally related genes. We demonstrated the method using three well-characterized mouse models of atopic dermatitis and showed that singIST could both reflect known disease biology and uncover new insights into disease mechanisms. Further, we developed statistical tests to validate the trained models. Our approach is flexible and broadly applicable, making it a valuable tool for improving the selection and understanding of disease models in biomedical research.

bioinformatics↗

bioRxiv2024

Smile-to-Bert: A BERT architecture trained for physicochemical properties prediction and SMILES embeddings generation

Molecular property prediction is crucial for drug discovery. Over the years, deep learning models have been widely used for these tasks; however, large datasets are often needed to achieve strong performances. Pre-training models on vast unlabeled data has emerged as a method to extract contextualized embeddings that boost performance on smaller datasets. The Simplified Molecular Input Line Entry System (SMILES) encodes molecular structures as strings, making them suitable for natural language processing. Transformers, known for capturing long-range dependencies, are well suited for processing SMILES. One such transformer-based architecture is Bidirectional Encoder Representations from Transformers (BERT), which only uses the encoder part of the Transformer and performs classification and regression tasks. Pre-trained transformer-based architectures using SMILES have significantly improved predictions on smaller datasets. Public data repositories such as PubChem, which provide SMILES, among other data, are essential for pre-training these models. SMILES embeddings that combine chemical structure and physicochemical property information could further improve performance on tasks such as Absorption, Distribution, Metabolism, Excretion, and Toxicity prediction. To this end, we introduce Smile-to-Bert, a pre-trained BERT architecture designed to predict 113 RDKit-computed molecular descriptors from PubChem SMILES. This model generates embeddings that integrate both molecular structure and physicochemical properties. We evaluate Smile-to-Bert on 22 datasets from the Therapeutics Data Commons and compare its performance with that of the 2-encoder model and a Transformer model. Smile-to-Bert achieves the best result on one dataset, while the combination of Smile-to-Bert with the other models leads to improved performance on 8 datasets. Additionally, the state-of-the-art Transformer is applied to Absorption, Distribution, Metabolism, Excretion, and Toxicity prediction for the first time, achieving the best performance on the Therapeutics Data Commons leaderboard of one dataset. Scientific Contribution: We present Smile-to-Bert, a model pre-trained to predict 113 molecular descriptors directly from SMILES representations. We evaluate our model on 22 property prediction datasets and compare the performance with other language models. We demonstrate that integrating our model with two other pre-trained models improves the performance in 8 datasets.

bioinformatics↗

bioRxiv2020

MultiPaths: a Python framework for analyzing multi-layer biological networks using diffusion algorithms

SummaryHigh-throughput screening yields vast amounts of biological data which can be highly challenging to interpret. In response, knowledge-driven approaches emerged as possible solutions to analyze large datasets by leveraging prior knowledge of biomolecular interactions represented in the form of biological networks. Nonetheless, given their size and complexity, their manual investigation quickly becomes impractical. Thus, computational approaches, such as diffusion algorithms, are often employed to interpret and contextualize the results of high-throughput experiments. Here, we present MultiPaths, a framework consisting of two independent Python packages for network analysis. While the first package, DiffuPy, comprises numerous state-of-the-art diffusion algorithms applicable to any generic network, the second, DiffuPath, enables the application of these algorithms on multi-layer biological networks. To facilitate its usability, the framework includes a command line interface, reproducible examples, and documentation. To demonstrate the framework, we conducted several diffusion experiments on three independent multi-omics datasets over disparate networks generated from pathway databases, thus, highlighting the ability of multi-layer networks to integrate multiple modalities. Finally, the results of these experiments demonstrate how the generation of harmonized networks from disparate databases can improve predictive performance with respect to individual resources. AvailabilityDiffuPy and DiffuPath are publicly available under the Apache License 2.0 at https://github.com/multipaths. Contactsergi.picart@upc.edu and daniel.domingo.fernandez@scai.fraunhofer.de

bioinformatics↗

bioRxiv2020

The effect of statistical normalisation on network propagation scores

MotivationNetwork diffusion and label propagation are fundamental tools in computational biology, with applications like gene-disease association, protein function prediction and module discovery. More recently, several publications have introduced a permutation analysis after the propagation process, due to concerns that network topology can bias diffusion scores. This opens the question of the statistical properties and the presence of bias of such diffusion processes in each of its applications. In this work, we characterised some common null models behind the permutation analysis and the statistical properties of the diffusion scores. We benchmarked seven diffusion scores on three case studies: synthetic signals on a yeast interactome, simulated differential gene expression on a protein-protein interaction network and prospective gene set prediction on another interaction network. For clarity, all the datasets were based on binary labels, but we also present theoretical results for quantitative labels. ResultsDiffusion scores starting from binary labels were affected by the label codification, and exhibited a problem-dependent topological bias that could be removed by the statistical normalisation. Parametric and non-parametric normalisation addressed both points by being codification-independent and by equalising the bias. We identified and quantified two sources of bias -mean value and variance- that yielded performance differences when normalising the scores. We provided closed formulae for both and showed how the null covariance is related to the spectral properties of the graph. Despite none of the proposed scores systematically outperformed the others, normalisation was preferred when the sought positive labels were not aligned with the bias. We conclude that the decision on bias removal should be problem and data-driven, i.e. based on a quantitative analysis of the bias and its relation to the positive entities. AvailabilityThe code is publicly available at https://github.com/b2slab/diffuBench Contactsergi.picart@upc.edu

bioinformatics↗

bioRxiv2018

Benchmarking network propagation methods for disease gene identification

BackgroundIn-silico identification of potential disease genes has become an essential aspect of drug target discovery. Recent studies suggest that one powerful way to identify successful targets is through the use of genetic and genomic information. Given a known disease gene, leveraging intermolecular connections via networks and pathways seems a natural way to identify other genes and proteins that are involved in similar biological processes, and that can therefore be analysed as additional targets.\n\nResultsHere, we systematically tested the ability of 12 varied network-based algorithms to identify target genes and cross-validated these using gene-disease data from Open Targets on 22 common diseases. We considered two biological networks, six performance metrics and compared two types of input gene-disease association scores. We also compared several cross-validation schemes and showed that different choices had a remarkable impact on the performance estimates. When seeding biological networks with known drug targets, we found that machine learning and diffusion-based methods are able to find novel targets, showing around 2-4 true hits in the top 20 suggestions. Seeding the networks with genes associated to disease by genetics resulted in poorer performance, below 1 true hit on average. We also observed that the use of a larger network, although noisier, improved overall performance.\n\nConclusionsWe conclude that machine learning and diffusion-based prioritisers are suited for drug discovery in practice and improve over simpler neighbour-voting methods. We also demonstrate the large effect of several factors on prediction performance, especially the validation strategy, input biological network, and definition of seed disease genes.

genetics↗