Bio-Marker Composition for Prediction of Drug Sensitivity Estimation Method for Prediction of Drug Sensitivity using Bio-Marker Composition and Diagnosing Chip for Detection of Bio-Marker Composition for Prediction of Drug Sensitivity

The present invention relates to genetic biomarker labeling scan (GBLscan) which is a drug indication and reaction prediction system and method as a novel learning model, capable of reliably predicting drug reactivity by binding analysis of a drug′s molecular profile after transforming specific gene...

Full description

Saved in:
Bibliographic Details
Main Authors LEE SUN HO, CHOI ICWON, JUNG, JONG SUN, HONG JONG HUI
Format Patent
LanguageEnglish
Korean
Published 19.05.2020
Subjects
Online AccessGet full text

Cover

Loading…
More Information
Summary:The present invention relates to genetic biomarker labeling scan (GBLscan) which is a drug indication and reaction prediction system and method as a novel learning model, capable of reliably predicting drug reactivity by binding analysis of a drug′s molecular profile after transforming specific genetic variation fingerprints related to diseases including cancer into a haplotype with functional information. The present invention comprises: a learning module that learns a reactivity correlation of constituent information constituting a drug to genetic information contained in a genome from collected learning information by linear regression and deep learning machine learning; a prediction module that receives analysis information and calculates a prediction result of drug reactivity to the genome included in the analysis information; and a storage module that stores a reactivity prediction algorithm learned by the learning module, wherein the learning information is drug reactivity information for in vitro cell line and in vivo clinical studies. According to the present invention, it is possible to predict the degree of drug reactivity with the genome whose pharmacological effect is unknown from the results of drug reactivity to the genome collected from clinical trials. 본 발명은 암을 포함하는 질병 관련 특이 유전자 변이 지문 (Genetic Variation Fingerprints)을 기능정보를 가진 하플로타입으로 변형 후 약물의 분자 프로파일의 결합분석에 의해 약물의 반응성을 신뢰성 있게 예측할 수 있는 새로운 학습모델인 약물 적응증 및 반응 예측 시스템 및 방법인 GBLscan(Genetic biomarker Labeling Scan)에 관한 것으로, 본 발명은 수집된 학습정보로부터 유전체에 포함된 유전정보에 대한 약물을 구성하는 구성정보의 반응성 상관관계를 선형 희귀 및 딥러닝 기계학습에 의해 학습하는 학습모듈과; 분석정보를 수신하여 상기 분석정보에 포함된 유전체에 대한 약물의 반응성 예측결과를 산출하는 예측모듈과; 상기 학습모듈에 의해 학습된 반응성 예측알고리즘을 저장하는 저장모듈을 포함하여 구성되고: 상기 학습정보는, 생체 외 세포주 및 생체 내 임상연구에 대한 약물의 반응성 정보이다. 이와 같은 본 발명에 의하면, 본 발명에서는 임상시험으로부터 수집되는 유전체에 대한 약물의 반응성 결과들로부터, 약리 효과가 밝혀지지 않은 유전체와 약물의 반응성 정도를 예측할 수 있는 효과가 있다.
Bibliography:Application Number: KR20190098492