Data-independent acquisition (DIA) is an emerging technology for quantitative proteomic analysis of large cohorts of samples. However, sample-specific spectral libraries built by data-dependent acquisition (DDA) experiments are required prior to DIA analysis, which is time-consuming and limits the identification/quantification by DIA to the peptides identified by DDA. Herein, we propose DeepDIA, a deep learning-based approach to generate in silico spectral libraries for DIA analysis. We demonstrate that the quality of in silico libraries predicted by instrument-specific models using DeepDIA is comparable to that of experimental libraries, and outperforms libraries generated by global models. With peptide detectability prediction, in silico libraries can be built directly from protein sequence databases. We further illustrate that DeepDIA can break through the limitation of DDA on peptide/protein detection, and enhance DIA analysis on human serum samples compared to the state-of-the-art protocol using a DDA library. We expect this work expanding the toolbox for DIA proteomics.

Background/Aim: Pancreatic cancer (PC) is currently the fourth leading cause of cancer-related mortality worldwide. Peripheral blood mononuclear cells (PBMCs) is a subpopulation of accessible and functional immune cells. Comparative analysis of the proteome of PBMCs can help us elucidate the mechanism of disease and find potential biomarkers for diagnosis. Materials and Methods: PBMCs were collected from healthy individuals, patients with benign diseases, and pancreatic cancer. iTRAQ-2DLC-MS/MS and SWATH methodologies were applied to make a comparative proteomics analysis of PBMCs. Results: A total of 3,357 proteins with a false discovery rate (FDR) <1% were identified, of which 114 proteins were found dysregulated in the PC group. An extensive SWATH library was constructed which showed a potential application for large scale clinical sample analysis. Conclusion: A PBMCs proteome with extensive protein representation was achieved, which will potentially allow the identification of novel biomarkers for PC.

The prevalence of intracranial aneurysm (IA) is increasing, and the consequences of its rupture are severe. This study aimed to reveal specific, sensitive, and non‐invasive biomarkers for diagnosis and classification of ruptured and unruptured IA, to benefit the development of novel treatment strategies and therapeutics altering the course of the disease. We first assembled an extensive candidate biomarker bank of IA, comprising up to 717 proteins, based on altered proteins discovered in the current tissue and serum proteomic analysis, as well as from previous studies. Mass spectrometry assays for hundreds of biomarkers were efficiently designed using our proposed deep learning‐based method, termed DeepPRM. A total of 113 potential markers were further quantitated in serum cohort I ( n = 212) & II ( n = 32). Combined with a machine‐learning‐based pipeline, we built two sets of biomarker combinations (P6 & P8) to accurately distinguish IA from healthy controls (accuracy: 87.50%) or classify IA rupture patients (accuracy: 91.67%) upon evaluation in the external validation set ( n = 32). This extensive circulating biomarker development study provides valuable knowledge about IA biomarkers.