Sira Sriswasdi

Mass Spectrometry-Based Profiling of Personalized Immunopeptidomes in Thai Renal Cell Carcinoma

This study profiles immunopeptidomes of Thai RCC patients, identifying tumor-specific neoantigens and non-canonical peptides, and establishes a framework for personalized cancer immunotherapy.

Connectome Principal Component Drives Cross-Dataset Replication and Clinical Prediction in Symptom Lesion Network Mapping

This study shows that symptom lesion network mapping (sLNM) results are biased toward the first principal component of the normative connectome, leading to spurious convergence and non-specific clinical predictions, which can be corrected by controlling for PC1.

SonicParanoid2: fast, accurate, and comprehensive orthology inference with machine learning and language models

SonicParanoid2 improves orthology inference speed and accuracy using machine learning and language models, achieving 18X faster performance and higher recall on benchmark datasets.

Label-free tumor cells classification using deep learning and high-content imaging

The paper develops a deep learning model to classify tumor cells from normal cells using high-content imaging, achieving an AUROC of 0.78 on a dataset of organoid-derived cells.

An explainable self-attention deep neural network for detecting mild cognitive impairment using multi-inputbdigital drawing tasks

This paper introduces an explainable self-attention deep neural network to detect mild cognitive impairment using multi-input digital drawing tasks, achieving better classification performance than existing models.

Uncovering thousands of new HLA antigens and phosphopeptides with deep learning-based sequence-mask-search de novo peptide sequencing framework

The paper introduces SMSNet, a deep learning framework for de novo peptide sequencing that identifies thousands of new HLA antigens and phosphopeptides with high accuracy.

MHCSeqNet: A deep neural network model for universal MHC binding prediction

MHCSeqNet is a deep learning model for predicting MHC binding affinity, outperforming existing methods and generalizing to new MHC alleles.

Supervised machine learning to decipher the complex associations between neuro-immune biomarkers and quality of life in schizophrenia

This study uses machine learning to identify predictors of health-related quality of life in schizophrenia, finding that neuro-immune biomarkers and cognitive impairments significantly impact HR-QoL.