Daniele Raimondi

pyreml is a GPU-accelerated Python library for linear mixed models that outperforms existing closed-source REML software in speed and functionality.

The paper shows that genome replication in yeast requires the cooperative activation of thousands of origins across the genome, driven by ORC binding and 3D chromosome folding.

This paper reports on the fourth CAFA challenge, showing improvements in Gene Ontology term prediction but highlighting challenges in new evaluation settings.

The paper identifies how batch correction methods can distort gene-gene correlations and introduces a metric to quantify this issue.

This paper shows that popular scRNA-seq batch correction methods fail to fully remove batch signal, which can affect downstream analyses, and introduces a supervised metric, BPS, to better assess correction effectiveness.

This paper identifies flaws in current Drug Response Prediction evaluation methods and proposes new strategies to improve their reliability and generalization.

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.

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

geneRNIB provides a benchmark for gene regulatory network inference, showing simpler models often outperform complex ones in perturbation-informed metrics.

MetDecode uses methylation patterns to estimate tumor proportion and identify tissue origin in cell-free DNA samples from cancer patients.

The paper defines six low energy conformational states for amino acid residues using solution NMR data and introduces a variability parameter to assess conformational state transitions.

This paper introduces an optimal transport-based domain adaptation method to correct technical biases in cfDNA sequencing data, improving cancer detection and enabling cohort integration.

The paper introduces DisoMine, a method predicting intrinsically disordered regions using protein dynamics and sequence context, outperforming existing predictors.

The paper evaluates 43 methods for predicting intrinsically disordered regions and binding regions in proteins, finding that deep learning methods outperform older physicochemical approaches but binding prediction remains challenging.