Biology subjects

De Moor, B.

Publications and source records attributed to De Moor, B..

9 recordsLinked to original sources

bioRxiv2026

AdventML: Advanced Enzyme Temperature Prediction with Transformer-Based Embeddings and Resampling Strategies

Accurate prediction of enzymes optimal catalytic temperature (Topt) is crucial in biotechnology, as enzymes with extreme Topt values are highly desirable for reactions at extreme temperatures and for their general stability. However, experimental determination of Topt is costly, labor-intensive, and time-consuming. Meanwhile, existing computational methods suffer from small and imbalanced datasets, subop-timal predictions at extreme temperatures, and insufficient validation. In this study, we address these challenges by expanding the Topt dataset and validating on an independent test set based on sequence similarity. We further tackle these limitations by comparing multiple resampling techniques to improve predictions at extremes and by considering diverse protein rep-resentations and multiple machine learning architectures. Overall, the best performing models reached R2 {approx} 0.64 with MAE {approx} 7-8 {degrees}C, while extreme resampling improved tail performance (reducing tail MAE by up to ~1.8 {degrees}C). Notably, our models show improved performance over state-of-the-art prediction models. We also demonstrate that accurate prediction of Topt is achievable even in the absence of organ-ism growth temperature (OGT). Our Topt prediction models are made freely available as AdventML on GitHub.

bioinformatics↗

bioRxiv2025

CLEAR-AF: Improved Autofluorescence Subtraction for Multiplexed Tissue Imaging via Polar Transformation and Gaussian Mixture Models

Cyclic Multiplexed ImmunoFluorescence (cMIF) can profile dozens to hundreds of proteins at subcellular resolution, offering unprecedented insights into tissue architecture in health and disease. However, measured signals are affected by tissue autofluorescence (AF), which can confound molecular quantifications at the single-cell level. Existing computational strategies to subtract AF are not sensitive nor specific enough and often leave residual background or over subtract weak markers. To address current limitations, we introduce CLEAR-AF (Coordinate-transformed Local Estimation and Adaptive Removal of AutoFluorescence), a subtraction framework where AF is calibrated for each acquired signal using a reference image. CLEAR-AF maps these intensities into polar coordinates to disentangle AF from true signal, and applies adaptive, distribution-aware thresholding to estimate and remove AF locally. Across multiple cMIF technologies, CLEAR-AF yields cleaner marker channels with improved specificity, greater sensitivity, and higher reproducibility relative to common AF removal approaches. By delivering more accurate per-cell measurements without workflow changes to acquisition, CLEAR-AF provides a practical, platform-agnostic step towards more reliable spatial proteomics at scale.

bioinformatics↗

bioRxiv2025

CONCLAVE: CONsensus CLustering with Annotation-Validation Extrapolation for cyclic multiplexed immunofluorescence data

High-dimensional cyclic multiplexed immunofluorescence (cMIF) enables single-cell phenotyping within intact tissues. Cell annotations rely on a multi-step pipeline involving normalization, sampling, dimensionality reduction, and clustering, but the absence of standardized benchmarks for method selection--especially at the clustering stage--leads to inconsistent and less reproducible phenotyping. To address this, we developed CONCLAVE, a consensus-clustering-based workflow that optimizes upstream steps and integrates results from multiple clustering algorithms retaining only those cell labels supported by at least two independent methods. Through in-silico simulations and real-world cMIF datasets, CONCLAVE consistently outperformed single-clustering-method approaches in accuracy, reproducibility, and robustness, with improvements becoming more evident when mapped within spatial tissue contexts. Additionally, CONCLAVE includes a scoring module that flags regions likely to contain unreliable or inconsistent data, facilitating targeted quality control. In summary, CONCLAVE offers a robust framework for cell annotation in cMIF datasets, enhancing the reliability of downstream spatial proteomics analyses. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=118 SRC="FIGDIR/small/688190v1_ufig1.gif" ALT="Figure 1"> View larger version (45K): org.highwire.dtl.DTLVardef@14ddfecorg.highwire.dtl.DTLVardef@1a83a17org.highwire.dtl.DTLVardef@17dfc74org.highwire.dtl.DTLVardef@493c07_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2025

Machine learning for optimal growth temperature prediction of prokaryotes using amino acid descriptors

MotivationThe optimal growth temperature (OGT) of organisms is valuable in bioprospecting enzymes that work under extreme conditions. Existing OGT prediction models achieve high accuracy, but mainly capture trends of overrepresented groups in the training set including organisms that thrive at moderate temperatures and those from well-described taxa. ResultsIn this study, we incorporated weighted scoring and phylogenetic splits to improve the generalizability of the prediction models. We first built a new growth temperature dataset comprising more than 21,000 species distributed over all three domains of life, with special attention to include OGT and extreme temperature data. We then trained machine learning models on the OGT data of 6,401 prokaryotes using proteome-averaged amino acid descriptors. The best-performing model was the multilayer perceptron with a cross-validated RMSE of 5.07{degrees}C ({+/-} 0.24) and an R2 of 0.89 ({+/-} 0.04). The most important proteome features were related to backbone flexibility, charged residues, as well as surface accessibility. Availability and ImplementationThe MLP model is integrated in the command line tool OGTFinder and available under MIT license at: https://github.com/SC-Git1/OGTFinder.

microbiology↗

bioRxiv2024

COLLAGE: COnsensus aLignment of muLtiplexing imAGEs

Multiplexed immunohistochemistry (mIHC) enables the high-dimensional single-cell interrogation of pathological tissue samples. mIHC is commonly based on the collection of high-resolution images from repeated staining cycles of the same tissue sample. Images of individual cycles typically consist of smaller tiles that need to be stitched into larger composite images, while images from serial rounds require alignment in a shared set of coordinates to enable pixel-perfect data integration. Current algorithms for stitching and registration require solving a single large puzzle consisting of billions of pixels making them computationally expensive but moreover forcing them to introduce errors to close the puzzle, which significantly impact the downstream results and the single-cell profiles. Here, we present the development and evaluation of COLLAGE (COnsensus aLignment of muLtiplexing imAGEs), an innovative stitching and registration method that leverages on the complementarity of these two steps in a divide and conquer approach: in contrast to other algorithms, COLLAGE breaks the process down into thousands of small puzzles, enabling extensive parallelisation and not forcing errors in its solution. Because COLLAGE also includes AlgnQC, a novel deep-learning-based evaluation metric of registration quality, the quality of the resulting image stacks is consistently maximised, while images with errors are flagged in an automated way. COLLAGE is available via www.disscovery.org. O_FIG O_LINKSMALLFIG WIDTH=148 HEIGHT=200 SRC="FIGDIR/small/603557v1_ufig1.gif" ALT="Figure 1"> View larger version (65K): org.highwire.dtl.DTLVardef@18d0f5borg.highwire.dtl.DTLVardef@1eb2756org.highwire.dtl.DTLVardef@163bb95org.highwire.dtl.DTLVardef@b04453_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2024

QUAL-IF-AI: Quality Control of Immunofluorescence Images using Artificial Intelligence

Fluorescent imaging has revolutionized biomedical research, enabling the study of intricate cellular processes. Multiplex immunofluorescent imaging has extended this capability, permitting the simultaneous detection of multiple markers within a single tissue section. However, these images are susceptible to a myriad of undesired artifacts, which compromise the accuracy of downstream analyses. Manual artifact removal is impractical given the large number of images generated in these experiments, necessitating automated solutions. Here, we present QUAL-IF-AI, a multi-step deep learning-based tool for automated artifact identification and management. We demonstrate the utility of QUAL-IF-AI in detecting four of the most common types of artifacts in fluorescent imaging: air bubbles, tissue folds, external artifacts, and out-of-focus areas. We show how QUAL-IF-AI outperforms state-of-the-art methodologies in a variety of multiplexing platforms achieving over 85% of classification accuracy and more than 0.6 Intersection over Union (IoU) across all artifact types. In summary, this work presents an automated, accessible, and reliable tool for artifact detection and management in fluorescent microscopy, facilitating precise analysis of multiplexed immunofluorescence images.

bioinformatics↗

bioRxiv2023

Integration of Multiple Spatial-Omics Modalities Reveals Unique Insights into Molecular Heterogeneity of Prostate Cancer

Recent advances in spatial omics methods are revolutionising biomedical research by enabling detailed molecular analyses of cells and their interactions in their native state. As most technologies capture only a specific type of molecules, there is an unmet need to enable integration of multiple spatial-omics datasets. This, however, presents several challenges as these analyses typically operate on separate tissue sections at disparate spatial resolutions. Here, we established a spatial multi-omics integration pipeline enabling co-registration and granularity matching, and applied it to integrate spatial transcriptomics, mass spectrometry-based lipidomics, single nucleus RNA-seq and histomorphological information from human prostate cancer patient samples. This approach revealed unique correlations between lipids and gene expression profiles that are linked to distinct cell populations and histopathological disease states and uncovered molecularly different subregions not discernible by morphology alone. By its ability to correlate datasets that span across the biomolecular and spatial scale, the application of this novel spatial multi-omics integration pipeline provides unprecedented insight into the intricate interplay between different classes of molecules in a tissue context. In addition, it has unique hypothesis-generating potential, and holds promise for applications in molecular pathology, biomarker and target discovery and other tissue-based research fields.

cancer biology↗

bioRxiv2020

Correspondence-aware manifold learning for microscopic and spatial omics imaging: a novel data fusion method bringing MSI to a cellular resolution.

High-dimensional molecular measurements are transforming the field of pathology into a data-driven discipline. While H&E stainings are still the gold standard to diagnose disease, the integration of microscopic and molecular information is becoming crucial to advance our understanding of tissue heterogeneity. To this end, we propose a data fusion method that integrates spatial omics and microscopic data obtained from the same tissue slide. Through correspondence-aware manifold learning, we can visualise the biological trends observed in the high-dimensional omics data at microscopic resolution. While data fusion enables the detection of elements that would not be detected taking into account the separate data modalities individually, out-of-sample prediction makes it possible to predict molecular trends outside of the measured tissue area. The proposed dimensionality reduction-based data fusion paradigm will therefore be helpful in deciphering molecular heterogeneity by bringing molecular measurements such as MSI to the cellular resolution.

bioinformatics↗

bioRxiv2020

Spatially-Aware Clustering of Ion Images in Mass Spectrometry Imaging Data Using Deep Learning

Computational analysis is crucial to capitalize on the wealth of spatio-molecular information generated by mass spectrometry imaging (MSI) experiments. Currently, the spatial information available in MSI data is often under-utilized, due to the challenges of in-depth spatial pattern extraction. The advent of deep learning has greatly facilitated such complex spatial analysis. In this work, we use a pre-trained neural network to extract high-level features from ion images in MSI data, and test whether this improves downstream data analysis. The resulting neural network interpretation of ion images, coined neural ion images, are used to cluster ion images based on spatial expressions. We evaluate the impact of neural ion images on two ion image clustering pipelines, namely DBSCAN clustering, combined with UMAP-based dimensionality reduction, and k-means clustering. In both pipelines, we compare regular and neural ion images from two different MSI datasets. All tested pipelines could extract underlying spatial patterns, but the neural network-based pipelines provided better assignment of ion images, with more fine-grained clusters, and greater consistency in the spatial structures assigned to individual clusters. Additionally, we introduce the Relative Isotope Ratio metric to quantitatively evaluate clustering quality. The resulting scores show that isotopical m/z values are more often clustered together in the neural network-based pipeline, indicating improved clustering outcomes. The usefulness of neural ion images extends beyond clustering towards a generic framework to incorporate spatial information into any MSI-focused machine learning pipeline, both supervised and unsupervised.

bioinformatics↗