Biology subjects

Huber, F.

Publications and source records attributed to Huber, F..

16 recordsLinked to original sources

bioRxiv2026

A Modular Chromosomally Integrated Yeast Surface Display Platform for Quantitative Serology and Immune Profiling

Precise quantification of antigen-specific antibodies is essential for assessing immune status and fundamental to both clinical and research settings. However, current serological platforms remain limited in their ability to combine quantitative performance, flexibility, and scalability. Here, we present a chromosomally integrated yeast surface display (YSD) system engineered for stable, quantitative, and scalable antibody profiling. Genomic integration of the YSD cassette supports uniform antigen presentation without continuous selective pressure, reducing variability associated with plasmid-based systems. Two distinct cell-wall anchoring strategies are implemented within a modular synthetic framework that enables rapid, cloning-free strain construction. Coupled with flow-cytometric readout, the system provides sensitive detection of antigen-specific antibodies with single-cell normalization for antigen expression. In addition, the combination of barcoded YSD strain libraries with fluorescence-activated cell sorting and next-generation sequencing enables multiplexed serological profiling across multiple samples and antigens. As a proof of concept, we demonstrate quantitative antibody detection against SARS-CoV-2 and extend the approach to hepatitis virus antigens, highlighting the utility of this platform as a flexible framework for serological assay development and immune profiling.

bioengineering↗

bioRxiv2026

Beyond the Default: Optimizing Molecular Networking with arteMIS

Metabolomics uses tandem mass spectrometry (MS/MS) data to gain structural insights of small molecules that play biological roles, generating datasets whose size and complexity demand systematic organisation. Molecular networking addresses this by representing MS/MS spectra as nodes and their pairwise similarity as edges, but its output is critically sensitive to user-defined parameters: similarity score cut-off, maximum component size, maximum links and minimum matching peaks. These parameters are routinely left at default values, which can either collapse interpretable molecular families into entangled "hairballs" or fragment them into disconnected singletons. In the absence of ground truth, no standardised framework exists to evaluate molecular networks or to assess whether their connections are robust to run-to-run variability present in metabolomic experiments. Here, we introduce arteMIS (Accelerated Ranking and Tuning using Multi-metric Interpretability across Scores), a framework for systematic parameter optimisation that uses Latin Hypercube Sampling to efficiently cover the four-dimensional parameter space and ranks candidate networks through a user-tuneable composite Z-score, combining topology- and chemistry-based metrics. This framework supports three complementary modes: global, seed, and target-class, adapting optimisation to fully unannotated datasets, curated subset of reference features or class-focused discovery, respectively. Benchmarking across four spectral libraries (~600 to ~13,000 spectra) and four scoring methods (Cosine, Modified Cosine, Spec2Vec, MS2DeepScore), we provide practical guidance for parameter selection as a function of scoring method and dataset size and show that optimal settings do not transfer between them. Top-ranked arteMIS configurations outperformed GNPS defaults in chemistry and topology metrics and produced networks with higher edge-stability under subsampling. Applied to actinobacteria and fungal samples, arteMIS rescued structurally meaningful families that remained fragmented under default settings. We conclude that arteMIS reframes molecular network construction from a default-driven step into a task-customisable optimisation.

bioinformatics↗

bioRxiv2025

Tissue specificity and chromosomal alterations shape divergent immune programs in HRD tumors

Homologous recombination deficiency (HRD) activates pro-inflammatory cGAS/STING signaling, positioning it as a biomarker for combining immune checkpoint blockade (ICB) and PARP inhibition (PARPi). However, the consequences of HRD on the immune landscape across cancers remain unclear. Here, we applied a pan-cancer HRD classifier to >10,000 tumors from The Cancer Genome Atlas and uncovered striking heterogeneity in immune activity. Compared to HR-proficient tumors, HRD tumors showed elevated inflammation in breast, ovarian, and endometrial cancers. These tumors exhibited robust activation of innate and adaptive immune pathways (IFN, NF-{kappa}B) and transcriptional hallmarks of senescence, angiogenesis, and adenosine signaling. In contrast, lung, head and neck, and melanoma HRD tumors displayed suppressed inflammation and evidence of immune escape through large-scale loss-of-heterozygosity (LOH) at IFNA/B, STING, and other loci. These tumors also frequently presented HLA LOH and oncogene amplifications, suggesting selection under immune pressure and replication stress. Together, our study resolves HRD tumors into two immune archetypes, immune-inflamed and immune-evasive, linked to chromosomal instability and lineage, informing biomarker-driven evaluation of immune checkpoint blockade/PARPi combinatorial therapies.

genomics↗

bioRxiv2025

Count your bits: more subtle similarity measures using larger radius count vectors

Quantifying molecular similarity is a cornerstone of cheminformatics, underpinning applications from virtual screening and nearest-neighbor search to chemical space visualization and the evaluation of machine-learning predictions. Although Tanimoto-comparisons of 2D fingerprints are widely used, the practical behavior of these similarity measures depends strongly on the fingerprint type, representation (binary vs. count), and whether fingerprints are folded into fixed-length vectors. Here, we systematically benchmark a broad set of common fingerprint types, including dictionary-based, circular (Morgan/FCFP), path-based (RDKit), topological-distance-based (Atom Pair), hybrid distance-encoded (MAP4), torsion, LINGO, and Avalon, across multiple large datasets and complementary evaluation tasks. We quantify fingerprint specificity via duplicate rates and mass discrepancies, characterize score distributions and compound-size dependence, assess top-k ranking agreement, and compare fingerprint similarities to a graph-based reference. Across benchmarks, count (and often log-count) variants generally improve specificity and structural alignment, while folding-induced bit collisions can strongly distort similarities for high-occupancy fingerprints, making unfolded variants particularly important for RDKit and often necessary for MAP4 on heterogeneous datasets. To support reproducible benchmarking and future extensions, we introduce the open-source Python library chemap, providing unified computation of folded, unfolded, and frequency-folded fingerprints and optimized similarity calculations. Scientific contributionWe introduce a multi-criteria benchmarking framework for molecular fingerprints that goes beyond classical virtual-screening retrieval tests by evaluating specificity (fingerprint duplicates), score behavior (including compound-size dependence), ranking agreement, and neighborhood structure on large, heterogeneous small-molecule datasets. Our results reveal that common default settings, especially folding for high-occupancy fingerprints, can introduce severe bit-collision artifacts, and that count (often log-count) and unfolded variants substantially improve specificity and agreement with structure-based references. We release the open-source Python library chemap to standardize these fingerprint variants and enable reproducible, extensible benchmarking for future fingerprint development.

bioinformatics↗

bioRxiv2024

Reliable cross-ion mode chemical similarity prediction between MS2 spectra

Mass spectrometry is a cornerstone of untargeted metabolomics, enabling the characterization of metabolites in both positive and negative ionization modes. However, comparisons across ionization modes have remained a substantial challenge due to the distinct fragmentation patterns produced by each polarity. To overcome this barrier, we present MS2DeepScore 2.0, a machine learning-based model to predict chemical similarity between mass fragmentation spectra, which works both between different and the same ionization modes. We demonstrate the utility of MS2DeepScore 2.0 in three case studies, where MS2DeepScore enabled cross-ionization mode molecular networking, enhancing data exploration and metabolite annotation. To ensure robustness, we have implemented a quality estimation method that flags spectra with low information content or those dissimilar to the training data, thereby minimizing false predictions. Altogether, MS2DeepScore 2.0 extends our current capabilities in organizing, exploring, and annotating untargeted metabolomics profiles.

bioinformatics↗

bioRxiv2024

Systematic mapping of antibiotic cross-resistance and collateral sensitivity with chemical genetics

By acquiring or evolving resistance to one antibiotic, bacteria can become resistant to a second one, due to shared underlying mechanisms. This is called cross-resistance (XR) and further limits therapeutic choices. The opposite scenario, in which initial resistance leads to sensitivity to a second antibiotic, is termed collateral sensitivity (CS) and can inform cycling or combinatorial treatments. Despite their clinical relevance, our current knowledge of such interactions is limited, mostly due to experimental constraints in their assessment and lack of understanding of the underlying mechanisms. To fill this gap, we used published chemical genetic data on the impact of all Escherichia coli non-essential genes on resistance/sensitivity to 40 antibiotics, and devised a metric that robustly discriminates between known XR and CS antibiotic interactions. This metric, based on chemical genetic profile (dis)similarity between two drugs, allowed us to infer 404 XR and 267 CS interactions, thereby expanding the number of known interactions by more than 3-fold - including reclassifying 116 previously reported interactions. We benchmarked our results by validating 55 out of 59 inferred interactions via experimental evolution. By identifying mutants driving XR and CS interactions in chemical genetics, we recapitulated known and uncovered previously unknown mechanisms, and demonstrated that a given drug pair can exhibit both interactions depending on the resistance mechanism. Finally, we applied CS drug pairs in combination to reduce antibiotic resistance development in vitro. Altogether, our approach provides a systematic framework to map XR/CS interactions and their mechanisms, paving the way for the development of rationally-designed antibiotic combination treatments.

systems biology↗

bioRxiv2023

Mass2SMILES: deep learning based fast prediction of structures and functional groups directly from high-resolution MS/MS spectra.

Modern mass spectrometry-based metabolomics generates vast amounts of mass spectral data as part of the chemical inventory of biospecimens. Annotation of the resulting MS/MS spectra remains a challenging task that mostly relies on database interrogations, in silico prediction and interpretation of diagnostic fragmentation schemes and/or expert knowledge-based manual interpretations. A key limitation is additionally that these approaches typically leave a vast proportion of the (bio)chemical space unannotated. Here we report a deep neural network method to predict chemical structures solely from high-resolution MS/MS spectra. This novel approach initially relies on the encoding of SMILES strings from chemical structures using a continuous chemical descriptor space that had been previously implemented for molecule design. The deep neural network was trained on 83,358 natural product-derived MS/MS spectra of the GNPS library and of the NIST HRMS database with addition of the calculated neutral losses for those spectra. After this training and parameter optimization phase, the deep neural network approach was then used to predict structures from MS/MS spectra not included in the training data-set. Our current version, implemented in the Python programming language, accurately predicted 7 structures from 744 validation structures and the following 14 structures had a Tanimoto similarity score above 0.9 when compared to the true structure. It was also able to correctly identify two structures from the CASMI 2022 international contest. On average the Tanimoto similarity is of 0.40 for data of the CASMI 2022 international contest and of 0.39 for the validation data-set. Finally, our deep neural network is also able to predict the number of 60 functional groups as well as the molecular formula of chemical structures and adduct type for the analyzed MS/MS spectra. Importantly, this deep neural network approach is extremely fast, in comparison to currently available methods, making it suitable to predict on regular computers structures for all substances within large metabolomics datasets.

bioinformatics↗

bioRxiv2022

MS2Query: Reliable and Scalable MS2 Mass Spectral-based Analogue Search

Metabolomics-driven discoveries of biological samples remain hampered by the grand challenge of metabolite annotation and identification. Only few metabolites have an annotated spectrum in spectral libraries; hence, searching only for exact library matches generally returns a few hits. An attractive alternative is searching for so-called analogues as a starting point for structural annotations; analogues are library molecules which are not exact matches, but display a high chemical similarity. However, current analogue search implementations are not yet very reliable and relatively slow. Here, we present MS2Query, a machine learning-based tool that integrates mass spectral embedding-based chemical similarity predictors (Spec2Vec and MS2Deepscore) as well as detected precursor masses to rank potential analogues and exact matches. Benchmarking MS2Query on reference mass spectra and experimental case studies demonstrates an improved reliability and scalability. Thereby, MS2Query offers exciting opportunities for further increasing the annotation rate of complex metabolite mixtures and for discovering new biology.

bioinformatics↗

bioRxiv2022

Comparison of Cosine, Modified Cosine, and Neutral Loss Based Spectral Alignment For Discovery of Structurally Related Molecules

Spectrum alignment of tandem mass spectrometry (MS/MS) data using the modified cosine similarity and subsequent visualization as molecular networks have been demonstrated to be a useful strategy to discover analogs of molecules from untargeted MS/MS-based metabolomics experiments. Recently, a neutral loss matching approach has been introduced as an alternative to MS/MS-based molecular networking, with an implied performance advantage in finding analogs that cannot be discovered using existing MS/MS spectrum alignment strategies. To comprehensively evaluate the scoring properties of neutral loss matching, the cosine similarity, and the modified cosine similarity, similarity measures of 955,228 peptide MS/MS spectrum pairs and 10 million small molecule MS/MS spectrum pairs were compared. This comparative analysis revealed that the modified cosine similarity outperformed neutral loss matching and the cosine similarity in all cases. The data further indicated that the performance of MS/MS spectrum alignment depends on the location and type of the modification, as well as the chemical compound class of fragmented molecules.

bioinformatics↗

bioRxiv2021

MEMO: Mass Spectrometry-based Sample Vectorization to Explore Chemodiverse Datasets

In natural products research, chemodiverse extracts coming from multiple organisms are explored for novel bioactive molecules, sometimes over extended periods. Samples are usually analyzed by liquid chromatography coupled with fragmentation mass spectrometry to acquire informative mass spectral ensembles. Such data is then exploited to establish relationships among analytes or samples (e.g. via molecular networking) and annotate metabolites. However, the comparison of samples profiled in different batches is challenging with current metabolomics methods. Indeed, the experimental variation - changes in chromatographical or mass spectrometric conditions - often hinders the direct comparison of the profiled samples. Here we introduce MEMO - MS2 BasEd SaMple VectOrization - a method allowing to cluster large amounts of chemodiverse samples based on their LC-MS/MS profiles in a retention time agnostic manner. This method is particularly suited for heterogeneous and chemodiverse sample sets. MEMO demonstrated similar clustering performance as state-of-the-art metrics taking into account fragmentation spectra. More importantly, such performance was achieved without the requirement of a prior feature alignment step and in a significantly shorter computational time. MEMO thus allows the comparison of vast ensembles of samples, even when analyzed over long periods of time, and on different chromatographic or mass spectrometry platforms. This new addition to the computational metabolomics toolbox should drastically expand the scope of large-scale comparative analysis. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=66 SRC="FIGDIR/small/474089v1_ufig1.gif" ALT="Figure 1"> View larger version (17K): org.highwire.dtl.DTLVardef@c540c0org.highwire.dtl.DTLVardef@1e07e7org.highwire.dtl.DTLVardef@198e030org.highwire.dtl.DTLVardef@1a6a15e_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2021

EB3-informed dynamics of the microtubule stabilizing cap during stalled growth

Microtubule stability is known to be governed by a stabilizing GTP/GDP-Pi cap, but the exact relation between growth velocity, GTP hydrolysis and catastrophes remains unclear. We investigate the dynamics of the stabilizing cap through in vitro reconstitution of microtubule dynamics in contact with micro-fabricated barriers, using the plus-end binding protein GFP-EB3 as a marker for the nucleotide state of the tip. The interaction of growing microtubules with steric objects is known to slow down microtubule growth and accelerate catastrophes. We show that the lifetime distributions of stalled microtubules, as well as the corresponding lifetime distributions of freely growing microtubules, can be fully described with a simple phenomenological 1D model based on noisy microtubule growth and a single EB3-dependent hydrolysis rate. This same model is furthermore capable of explaining both the previously reported mild catastrophe dependence on microtubule growth rates and the catastrophe statistics during tubulin washout experiments.

cell biology↗

bioRxiv2021

MS2DeepScore - a novel deep learning similarity measure for mass fragmentation spectrum comparisons

Mass spectrometry data is one of the key sources of information in many workflows in medicine and across the life sciences. Mass fragmentation spectra are considered characteristic signatures of the chemical compound they originate from, yet the chemical structure itself usually cannot be easily deduced from the spectrum. Often, spectral similarity measures are used as a proxy for structural similarity but this approach is strongly limited by a generally poor correlation between both metrics. Here, we propose MS2DeepScore: a novel Siamese neural network to predict the structural similarity between two chemical structures solely based on their MS/MS fragmentation spectra. Using a cleaned dataset of >100,000 mass spectra of about 15,000 unique known compounds, MS2DeepScore learns to predict structural similarity scores for spectrum pairs with high accuracy. In addition, sampling different model varieties through Monte-Carlo Dropout is used to further improve the predictions and assess the models prediction uncertainty. On 3,600 spectra of 500 unseen compounds, MS2DeepScore is able to identify highly-reliable structural matches and predicts Tanimoto scores with a root mean squared error of about 0.15. The prediction uncertainty estimate can be used to select a subset of predictions with a root mean squared error of about 0.1. We demonstrate that MS2DeepScore outperforms classical spectral similarity measures in retrieving chemically related compound pairs from large mass spectral datasets, thereby illustrating its potential for spectral library matching. Finally, MS2DeepScore can also be used to create chemically meaningful mass spectral embeddings that could be used to cluster large numbers of spectra. Added to the recently introduced unsupervised Spec2Vec metric, we believe that machine learning-supported mass spectral similarity metrics have great potential for a range of metabolomics data processing pipelines.

bioinformatics↗

bioRxiv2020

Spec2Vec: Improved mass spectral similarity scoring through learning of structural relationships

Spectral similarity is used as a proxy for structural similarity in many tandem mass spectrometry (MS/MS) based metabolomics analyses such as library matching and molecular networking. Although weaknesses in the relationship between spectral similarity scores and the true structural similarities have been described, little development of alternative scores has been undertaken. Here, we introduce Spec2Vec, a novel spectral similarity score inspired by a natural language processing algorithm -- Word2Vec. Spec2Vec learns fragmental relationships within a large set of spectral data to derive abstract spectral embeddings that can be used to assess spectral similarities. Using data derived from GNPS MS/MS libraries including spectra for nearly 13,000 unique molecules, we show how Spec2Vec scores correlate better with structural similarity than cosine-based scores. We demonstrate the advantages of Spec2Vec in library matching and molecular networking. Spec2Vec is computationally more scalable allowing structural analogue searches in large databases within seconds.

bioinformatics↗

bioRxiv2020

matchms - processing and similarity evaluation of mass spectrometry data

Mass spectrometry data is at the heart of numerable applications in the biomedical and life sciences. With growing use of high throughput techniques researchers need to analyse larger and more complex datasets. In particular through joint effort in the research community, fragmentation mass spectrometry datasets are growing in size and number. Platforms such as MassBank (Horai et al. 2010), GNPS (Wang et al. 2016) or MetaboLights (Haug et al. 2020) serve as an open-access hub for sharing of raw, processed, or annotated fragmentation mass spectrometry data (MS/MS). Without suitable tools, however, exploitation of such datasets remains overly challenging. In particular, large collected datasets contain data aquired using different instruments and measurement conditions, and can further contain a significant fraction of inconsistent, wrongly labeled, or incorrect metadata (annotations).

bioinformatics↗

bioRxiv2019

Integrated Proteogenomic Deep Sequencing and Analytics Accurately Identify Non-Canonical Peptides in Tumor Immunopeptidomes

Efforts to precisely identify tumor human leukocyte antigen (HLA) bound peptides capable of mediating T cell-based tumor rejection still face important challenges. Recent studies suggest that non-canonical tumor-specific HLA peptides that derive from annotated non-coding regions could elicit anti-tumor immune responses. However, sensitive and accurate mass-spectrometry (MS)-based proteogenomics approaches are required to robustly identify these non-canonical peptides. We present an MS-based analytical approach that characterizes the non-canonical tumor HLA peptide repertoire, by incorporating whole exome sequencing, bulk and single cell transcriptomics, ribosome profiling, and a combination of two MS/MS search tools. This approach results in the accurate identification of hundreds of shared and tumor-specific non-canonical HLA peptides and of an immunogenic peptide from a downstream reading frame in the melanoma stem cell marker gene ABCB5. It holds great promise for the discovery of novel cancer antigens for cancer immunotherapy.

cancer biology↗

bioRxiv2017

A genome-wide resource for high-throughput genomic tagging of yeast ORFs

Here we describe a C-SWAT library for high-throughput tagging of Saccharomyces cerevisiae ORFs. It consists of 5661 strains with an acceptor module inserted after each ORF, which can be efficiently replaced with tags or regulatory elements. We validate the library with targeted sequencing and demonstrate its use by tagging the yeast proteome with bright fluorescent proteins, determining how sequences downstream of ORFs influence protein expression and localizing previously undetected proteins.

genomics↗