Biology subjects

Kaplan, G.

Publications and source records attributed to Kaplan, G..

6 recordsLinked to original sources

bioRxiv2026

Predicting Antibody Self-Association with Sequence Structure Fusion Models: The Central Role of CSI-BLI in Early Developability Screening

Antibody-based biologics are expanding rapidly, yet challenges in development from self-association, high viscosity, aggregation, and unfavorable clearance underscore the need for accurate in silico screening. Clone self-interaction biolayer interferometry (CSI-BLI) is a plate-based, low-material assay of weak, reversible self-association that serves as an early proxy for high-concentration viscosity and a complementary predictor of in vivo clearance. In a 246-mAb panel, CSI-BLI moderately correlates with viscosity; further, in hFcRn Tg32 mice (41 antibodies), CSI-BLI strongly associates with clearance. Here, we present an end-to-end framework that distinguishes high versus low self-interacting clones (CSI-BLI class) by coupling a fine-tuned protein language model (ESM-2) with residue-aligned 3D context from AlphaFold-predicted structures encoded as residue graphs. Disentangled multi-stream attention fuses sequence content, chain-aware positional information, and structural signals to capture spatially proximate interactions that are distant in sequence. Edit-distance-controlled splits across 1499 IgGs and 988 VHHs assess generalization. The structure-aware model achieves the highest hold-out performance (VHH-Fc F1 = 0.76; IgG F1 = 0.57), while a sequence-only disentangled variant outperforms a standard PLM baseline without structural inputs. Complementary biophysical feature-based models, built from AlphaFold structures and sequence/structure-derived physicochemical descriptors with cluster-aware selection, deliver robust, interpretable performance (VHH F1 = 0.72; IgG F1 = 0.57), with SHAP analyses highlighting charge/dipole, hydrophobicity, and aggregation-propensity drivers across CDRs and frameworks. This interaction-aware sequence-structure framework, supported by interpretable feature models, is extensible to other developability endpoints and broader protein classification tasks where joint modeling of language-derived representations and residue-level geometry is advantageous.

bioinformatics↗

bioRxiv2025

Accelerating Antibody Development: Sequence and Structure-Based Models for Predicting Developability Properties through Size Exclusion Chromatography

Experimental screening for biopharmaceutical developability properties typically relies on resource-intensive, and time-consuming assays such as size exclusion chromatography (SEC). This study highlights the potential of in silico models to accelerate the screening process by exploring sequence and structure-based machine learning techniques. Specifically, we compared surrogate models based on pre-computed features extracted from sequence and predicted structure with sequence-based approaches using protein language models (PLMs) like ESM-2. In addition to different end-to-end fine-tuning strategies for PLM, we have also investigated the integration of the structural information of the antibodies into the prediction pipeline through graph neural networks (GNN). We applied these different methods for predicting protein aggregation propensity using a dataset of approximately 1200 Immunoglobulin G (IgG1) molecules. Through this empirical evaluation, our study identifies the most effective in silico approach for predicting developability properties for SEC assays, thereby adding insights to existing screening efforts for accelerating the antibody development process.

bioinformatics↗

bioRxiv2024

Leveraging remote sensing to distinguish closely related beech species in assisted gene flow scenarios

European beech (Fagus sylvatica L.) forests are suffering under increasingly severe and frequent drought. Three closely related, hybridizing beech species, ranging from Bulgaria through Asia Minor and the Caucasus to Iran, offer potential resources for assisted gene flow (AGF) with the aim of increasing the adaptive capacity of European beech forests. However, due to similar morphology and leaf color, as well as hybridization, it is challenging to track the fate of introduced beech genotypes from these related species. Traditional identification methods relying on detailed morphological characterization and genetic testing are labor-intensive and costly, making them impractical for large-scale applications. Using multispectral data from PlanetScope SuperDove, we developed a classification approach that captures phenological differences between the European beech F. sylvatica and co-planted Caucasian beech (Fagus hohenackeriana Palibin). The approach focuses on key temporal windows and spectral features to optimize classification performance. We evaluated various machine learning algorithms with stratified spatial and temporal cross-validation on data from more than 200 genetically classified individuals in two well-studied sites in France and Switzerland, where Caucasian beech was introduced over a century ago. Our approach was then tested on three different study areas in Germany, where Caucasian beech was also planted, but without specific tree coordinates. Our results reveal consistent temporal and spectral differences during spring and autumn, aligning with budbreak and senescence periods. Most algorithms achieved classification accuracies of 90% and above. The algorithms effectively identified candidate zones for Caucasian beech within or near areas indicated by local foresters. This study demonstrates the potential of high-resolution multispectral satellite imagery and machine learning for scalable classification of closely related and hybridizing species, thereby facilitating forest management in the face of global change. HighlightsO_LIRemote sensing of unique phenological differences between closely related beech species C_LIO_LIModels with optimized classification parameters differentiated highly similar, introduced trees C_LIO_LIReal-world application to assisted gene flow sites yielded high performance and testable predictions C_LI

ecology↗

bioRxiv2023

Enhancement of antibody thermostability and affinity by computational design in the absence of antigen

Over the last two decades, therapeutic antibodies have emerged as a rapidly expanding domain within the field biologics. In silico tools that can streamline the process of antibody discovery and optimization are critical to support a pipeline that is growing more numerous and complex every year. In this study, DeepAb, a deep learning model for predicting antibody Fv structure directly from sequence, was used to design 200 potentially stabilized variants of an anti-hen egg lysozyme (HEL) antibody. We sought to determine whether DeepAb can enhance the stability of these antibody variants without relying on or predicting the antibody-antigen interface, and whether this stabilization could increase antibody affinity without impacting their developability profile. The 200 variants were produced through a robust highthroughput method and tested for thermal and colloidal stability (Tonset, Tm, Tagg), affinity (KD) relative to the parental antibody, and for developability parameters (non-specific binding, aggregation propensity, self-association). In the designed clones, 91% and 94% exhibited increased thermal and colloidal stability and affinity, respectively. Of these, 10% showed a significantly increased affinity for HEL (5-to 21-fold increase), with most clones retaining the favorable developability profile of the parental antibody. These data open the possibility of in silico antibody stabilization and affinity maturation without the need to predict the antibody-antigen interface, which is notoriously difficult in the absence of crystal structures.

molecular biology↗

bioRxiv2021

Aggregation state of Mycobacterium tuberculosis impacts host immunity and augments pulmonary disease pathology

Phagocytosis of Mycobacterium tuberculosis (Mtb) aggregates, rather than similar numbers of single bacilli, induces host macrophage death and favors bacterial growth. Here, we examined whether aggregation contributes to enhanced Mtb pathogenicity in vivo in rabbit lungs. Rabbits were exposed to infectious aerosols containing mainly Mtb-aggregates (Mtb-AG) or Mtb-single cells (Mtb-SC). The lung bacterial load, histology, and immune cell composition were investigated over time. Genome-wide transcriptome analysis, cellular and tissue-level assays, and immunofluorescent imaging were performed on lung tissue to define and compare differential immune activation and pathogenesis between Mtb-AG and Mtb-SC infection. Lung bacillary loads, disease scores, lesion size, and structure were significantly higher in Mtb-AG than in Mtb-SC infected animals. A differential immune cell distribution and activation were noted in the lungs and spleen of the two groups of infected animals. Mtb-AG infected animals also showed early induction of inflammatory network genes associated with necrosis and reduced host cell viability. Consistently, larger lung granulomas with clumped Mtb, extensive necrotic foci, and elevated matrix metalloproteases expression were observed in Mtb-AG infected rabbits. Our findings suggest that bacillary aggregation increases Mtb fitness for improved growth and accelerated lung inflammation and cell death, thereby exacerbating disease pathology in the lungs.

microbiology↗

bioRxiv2021

Aggregated Mycobacterium tuberculosis enhances the inflammatory response

Mycobacterium tuberculosis (Mtb) readily aggregates in culture and Mtb aggregates in the lung were observed in experimental Mtb infection. However, the physiological consequences of Mtb aggregation are incompletely understood. Here we examined the human macrophage transcriptional response to aggregated Mtb relative to infection with non-aggregated single or multiple bacilli per host cell. Infection with aggregated Mtb led to an early upregulation of pro-inflammatory associated genes and enhanced TNF signaling via the NF{kappa}B pathway. Both these pathways were significantly upregulated relative to infection with single bacilli, and TNF signaling was also significantly elevated relative to infection with multiple non-aggregated Mtb. Secretion of TNF and downstream cytokines were also enhanced. On a longer timescale, aggregate infection led to overall increased acidification per macrophage and a high proportion of death in these cells after aggregate phagocytosis. Host cell death did not occur when Mtb aggregates were heat killed despite such clumps being readily picked up. To validate that Mtb aggregates do occur in the human lung, we document Mtb aggregates surrounding a cavity in a human TB lesion. Aggregates may therefore be present in some lesions and elicit a stronger inflammatory response resulting in recruitment of additional phagocytes and their subsequent death, potentially leading to necrosis and transmission.

microbiology↗