Γεώργιος Παλιούρας

This paper uses generative and symbolic AI to forecast kidney cancer progression from cross-sectional data, providing interpretable models and molecular insights.

The paper introduces CPP2Vec, a Word2Vec-based method for predicting cell-penetrating peptides, and compares it with pre-trained LLM-based models, showing strong predictive performance.

KG-Bench benchmarks graph neural networks for drug repurposing using the Open Targets dataset, finding TransformerConv achieves the highest performance.

This study compares machine learning methods for classifying multiple sclerosis and Alzheimer's disease using genomic data from UK Biobank, finding logistic regression is most robust while deep learning shows high variability.

The paper introduces BioASQ-QA, a biomedical QA dataset with structured and unstructured data, including exact and ideal answers for multi-document summarization and IR experiments.

CRISPRedict provides accurate and interpretable predictions for CRISPR-Cas9 guide RNA efficiency using a simple linear model, outperforming some existing tools and showing robust performance across different datasets.

The paper uses machine learning to analyze genotyping data and identifies interactions between SNVs in late onset Alzheimer's disease, highlighting specific genomic profiles and interactions involving APOE and other genes.