Biology subjects

Söding, J.

Publications and source records attributed to Söding, J..

3 recordsLinked to original sources

bioRxiv2025

LoL-align: sensitive and fast probabilistic protein structure alignment

The ubiquitous availability of protein structures permits replacing sequence alignment with more accurate and sensitive structure alignment algorithms. LoL-align maximizes a local log-odds score for proteins to be homologous, given their intra-protein C - C distances. LoL-align is markedly more sensitive in detecting remote homologs than TMalign and DALI on single- and multi-domain benchmarks and achieves better alignment quality at 5-20x their speed. LoL-align is fully integrated into Foldseek.

bioinformatics↗

bioRxiv2025

SoftAlign: End-to-end protein structures alignment

With the recent breakthrough of highly accurate structure prediction methods, there has been a rapid growth of available protein structures. Efficient methods are needed to infer structural similarity within these datasets. We present an end-to-end alignment method, called SoftAlign, which takes as input the 3D coordinates of a protein pair and outputs a structural alignment. In addition to the traditional Smith-Waterman alignment method, we introduce a modified softmax alignment that shows very promising results for structure similarity detection. We demonstrate that the SoftAlign model is able to recapitulate TM-align alignments while running faster, and it is more accurate than Foldseek on alignment and classification tasks. Although SoftAlign is not the fastest method available, it is highly precise and can be used effectively with other prefilters. In addition to developing an end-to-end structural aligner, our main contribution is the introduction and analysis of a pseudo-alignment method based on softmax, which can be used with other architectures, even those not based on structural information. The code for SoftAlign is available at https://github.com/jtrinquier/SoftAlign.

bioinformatics↗

bioRxiv2021

Cooperativity boosts affinity and specificity of proteins with multiple RNA-binding domains

Numerous cellular processes rely on the binding of proteins with high affinity to specific sets of RNAs. Yet most RNA binding domains display low specificity and affinity, to the extent that for most RNA-binding domains, the enrichment of the best binding motif measured by high-throughput RNA SELEX or RNA bind-n-seq is usually below 10-fold, dramatically lower than that of DNA-binding domains. Here, we develop a thermodynamic model to predict the binding affinity for proteins with any number of RNA-binding domains given the affinities of their isolated domains. For the four proteins in which affinities for individual domains have been measured the model predictions are in good agreement with experimental values. The model gives insight into how proteins with multiple RNA-binding domains can reach affinities and specificities orders of magnitude higher than their individual domains. Our results contribute towards resolving the conundrum of missing specificity and affinity of RNA binding proteins and underscore the need for bioinformatic methods that can learn models for multi-domain RNA binding proteins from high-throughput in-vitro and in-vivo experiments.

molecular biology↗