Weberling, F.-M. | Search bioRxiv

bioRxiv2026

Deep generative embeddings of gene expression and splicing reposition the interpretation of single-cell transcriptomic signatures

Single-cell transcriptomic analysis predominantly derives cell identity from gene expression analysis, while alternative splicing is processed separately despite its fundamental role for cell homeostasis. To overcome the limits of separate investigations, we developed a probabilistic deep learning framework, Crecerelle, enabling resolution of the contributions of gene expression and alternative splicing in each cell. Crecerelle learns cell embeddings from gene expressions and alternative splicing isoforms, to decipher their mutually dependent impact on the functional characterisation of cells in a data-driven manner, exemplified for the Tabula Muris dataset. This is enabled through a zero-and-N-inflated Dirichlet-Multinomial for a variational autoencoder that learns cell embeddings solely from splicing profiles, as well as a bi-modal variational autoencoder with a relevance-weighted mixture-of-experts variational posterior to consolidate the modality-specific contribution at single-cell level. Crecerelle reveals cell-type-specific isoform markers as well as subpopulations with unique isoforms and uncovers regulatory and disease-associated pathways not detected by gene expression analyses alone. This scalable and interpretable framework thus allows a more holistic study of transcriptomic regulation and will open a route to modality-relevance-weighted investigations across single-cell multiomics datasets and their influence on cellular homeostasis, tissue development and disease phenotypes.

cell biology↗

bioRxiv2024

Microdroplet screening rapidly profiles a biocatalyst to enable its AI-assisted engineering

Engineering enzymes for increased efficiency is key to enabling sustainable, green biocatalytic production processes in the chemical and pharmaceutical industries. This challenge can be tackled from two angles: by directed evolution, based on labour-intensive experimental testing of enzyme variant libraries, or by computational methods, where data-dependent algorithms relating sequence and function are used to predict biocatalyst improvements. Here, we combine both approaches into a two-week, low-cost workflow, in which ultra-high throughput screening of a library of imine reductases (IREDs) in microfluidic devices provides not only selected hits, but also long-read sequence data linked to fitness scores of >17 thousand enzyme variants. We demonstrate the engineering of an IRED for chiral amine synthesis by mapping its local fitness landscape in one go, ready to be used for interpretation and extrapolation by protein engineers with the help of machine learning (ML). We calculate position-dependent mutability and combinability scores of mutations and comprehensively illuminate a complex interplay of mutations driven by synergistic, often positively epistatic effects. When interpreted by easy-to-use regression and tree-based ML algorithms designed for random whole-gene mutagenesis data, 3-fold improved hits initially obtained from experimental screening are extrapolated further to give another order of magnitude improvement (23-fold in kcat) after testing only a handful of designed mutants. Predictions succeed in >80% of cases. The catalytic features discovered in one IRED are shown to be portable and confer activity on IREDs with [~]50% homology. Our campaigns yield biocatalytically efficient IREDs and are paradigmatic for future enzyme engineering efforts that rely on large sequence-function maps, profiling how a biocatalyst responds to mutation. In the age of predictive biology, these maps will chart the way to improved function by exploiting the synergy of rapid experimental screening combined with ML evaluation and extrapolation.

synthetic biology↗