Castrense Savojardo

Identification of potential colorectal-cancer-protective bioactive natural compounds in the human gut microbiome

The paper identifies four BGCs enriched in healthy individuals and one linked to a curacin-A-like metabolite, suggesting potential CRC-protective compounds from the gut microbiome.

Acinetobacter enrichment shapes composition and function of the bacterial microbiota of field-grown tomato plants

The paper identifies Acinetobacter sp. as a dominant bacterium in tomato plant microbiota and shows its functional role in nutrient mobilization and stress adaptation, while decoupling its presence from antimicrobial resistance genes.

Evaluation of enzyme activity predictions for variants of unknown significance in Arylsulfatase A

The paper evaluates variant effect predictions for ARSA gene variants, finding that a bootcamp-trained model outperformed others and that deep learning slightly improved performance over simpler methods.

Critical assessment of missense variant effect predictors on disease-relevant variant data

The paper evaluates missense variant effect predictors using CAGI6 data, finding strong performance across settings and highlighting biases in some predictors.

DDGemb: predicting protein stability change upon single- and multi-point variations with embeddings and deep learning

DDGemb predicts protein stability changes from single- and multi-point variations using embeddings and deep learning, achieving state-of-the-art performance.

Critical assessment of variant prioritization methods for rare disease diagnosis within the Rare Genomes Project

The paper evaluates variant prioritization methods for rare disease diagnosis using data from the Rare Genomes Project, finding that top models identify causal variants in most solved families and uncover potential candidates in unsolved ones.

CoCoNat: a novel method based on deep-learning for coiled-coil prediction

CoCoNat predicts coiled-coil boundaries, register annotations, and oligomerization states using deep learning, outperforming existing methods on multiple metrics.

ISPRED-SEQ: Deep neural networks and embeddings for predicting interaction sites in protein sequences.

ISPRED-SEQ predicts protein-protein interaction sites from sequences using deep neural networks and outperforms existing methods.

E-SNPs&GO: Embedding of protein sequence and function improves the annotation of pathogenic variants.

E-SNPs&GO improves pathogenic variant annotation by embedding protein sequences and function, achieving a MCC score of 0.71 on a blind test set.

The CAFA challenge reports improved protein function prediction and new functional annotations for hundreds of genes through experimental screens

The paper reports results from the CAFA3 challenge, showing improved computational protein function prediction and new annotations for over 1000 genes through experimental screens in C. albicans, P. aureginosa, and Drosophila.