Design And Implementation Of A Scalable Tumor Precision Medicine Database And Searching System

Increasingly affordable and available high-throughput techniques and rapidly improving understanding of clinically relevant genomic variants are major drivers for precision oncology.To implement genomically informed cancer therapy,clinicians need access to the state-of-the-art and comprehensive variant information including variant annotation,mutational effect at functional level,available treatment options and potential clinical trials.However,in practice,the tasks of collecting such dispersed,heterogeneous,and even conflicting,varying medical knowledge data and then matching them with patient variants are rather challenging.We thus designed and constructed a scalable tumor precision medicine database and searching system.It can identify multiple molecular features from sequencing data of patient samples,including SNPs/Indels,gene fusions,expression alterations,alternative splicing,etc.Then it can match these patient variants with relevant medical knowledge according to specific rules and integrate them into a report for clinical decision support.Our system consists of a database system and an information searching system.The database system mainly stores data of patient omics data,precision medicine knowledge and reference information.The searching system contains a matching algorithm and a reporting tool,which respectively match patient omics data with medical actions(e.g.,therapy and dosage recommendations,available clinical trials)and automatically generate evidence-based clinical reports.The system takes advantage of the three-tier architecture,mutilayered concept model and a specific expressive and query language to be highly extensible,modular and loosely coupled in dealing with constantly updating and growing medical data and quickly evolving analytical approaches.