Clinical trialObservationalCompleted
Validation of Useful Markers Generated by Next Generation Bio-data Based Genome Research and Cohort Study
Multiple Biomarker Development Through Validation of Useful Markers Generated by Next Generation Bio-data Based Genome Research and Cohort Study
- Outcome
- Finished, no results posted
- Phase
- Observational
- Status
- Completed
- Study type
- Observational
- Start
- September 2016
- Ended
- September 2016
- Enrollment
- 232 participants
- Sites
- 1
- Primary outcomes
- AUC(area under curve)
- Sponsor type
- Industry
- Results posted
- No
- Last update
- 31 January 2020
These fields come from the trial’s registry entry. A summary of its population, design, endpoints and results, each tied to a quote, has not been written for this trial yet.
Ask this document
Answers come only from this document’s own text, as highlighted quotes. If it doesn’t say, you are told so instead of given a guess.
Try
Source text
Brief summary public domain (ClinicalTrials.gov)
Multiple biomarker development through validation of useful markers generated by next generation bio-data based genome research and cohort study
Detailed description public domain (ClinicalTrials.gov)
1. Objectives The study will be performed to develop the integrated analytical methods of genomic data and clinical data and the bio-control network analysis, through which knowledge-based integrated analysis system can be developed and then biomarker for early diagnosis and treatment of pancreatic cancer and bile duct cancer, and finally the customized disease management system. Also, it is to confirm the effectiveness of diagnostic chip for research purpose by applying pancreatic/bile duct cancer-specific biomarker, miRNA, found through the integrated analysis of genomic data and clinical data of patients with pancreatic/bile duct cancer to the blood of patients with pancreatic/bile duct cancer. 2. Effective evaluation method The discrimination and calibration for algorithm through the diagnostic chip of each cancer type will all be examined using 10-fold cross-validation (100 repetitions). In the 10-fold cross-validation, the data is randomly divided into 10 same sized data, among which 9 are used in making a model for training and the remaining 1 is applied for test, and this process is randomly and independently repeated for 100 times. The 10-fold cross-validated AUC is calculated to see the discrimination of diagnostic chip of each cancer type, and the 95% confidence interval is presented by non-parametric method. The 10-fold cross-validated calibration plot is presented to see the calibration of diagnostic chip of each cancer type. The calibration plot visually demonstrate the degree of prediction by comparing the prediction probability of each group and the ratio of actual cancer patients after listing the prediction probability in the order and dividing it with regular intervals. Then, for the same subjects, the AUC of the CA 19-9, the existing cancer diagnostic tool, is calculated and the 95% confidence interval is presented. To compare the diagnostic chip of each cancer type and the AUC of CA 19-9, p-value is calculated by non-parametric method of 10-fold cross-validated AUC.