Supplementing the existing collection with approximately 2,600 new sets of multi-omic sequencing data, including germline and somatic exomes (WES), tumor and blood transcriptomes (RNA-Seq), tumor and liquid biopsy microtranscriptomes (miRNA: small RNA-Seq and nCounter), and tumor methylomes (RRBS). Combined with archival data (approximately 3,400 sets), this will create a database of over 6,000 multi-omic profiles.

Preparation of complete, processed sequencing data through quality control, filtering, and standardization of bioinformatic pipelines and data formats according to SOPs, enabling their use in advanced predictive models.

Integration of multiple omics layers (somatic and germline WES, tumor and blood RNA-Seq, tumor small RNA-Seq, tumor RRBS, and miRNA nCounter – liquid biopsy) with clinical data, followed by trans-omic analyses to identify candidate biomarkers for carcinogenesis processes and disease progression prediction.

Identification of key molecular features using dimensionality reduction methods, statistical selection, and machine learning algorithms, combined with expert knowledge, to pinpoint biomarkers with the highest diagnostic and prognostic potential.

Development of models integrating multi-omic and clinical data, along with their validation for predictive performance and clinical utility. Examples of planned evaluations include: a non-invasive molecular diagnostic test based on a blood sample to detect an ongoing cancer of a specific type or stage, and a prognostic test predicting a high risk of early death following tumor resection surgery.

Conducting qPCR analyses for selected biomarkers to confirm their clinical utility, validating models using qPCR data, and preparing prototype tests suitable for implementation in routine diagnostics.