This program will perform summary statistics imputation for one phenotype that is hard-to-collect using a set of easy-to-collect phenotypes.
Example of usage: python imputeStatFromSummary.py -i YOURPATH/inputHaemRBC/ -n HaemGenRBC -f Logferr -o test -r correlation.data
This command will try to impute the Logferr level using the corretion between all the phenotypes.
- It assumes that correlation.data is a matrix with column header name. For example:
| Ferr | Logferr | HB | HCT | MCH |
|---|---|---|---|---|
| 1 | 0.81605436 | 0.24744599 | 0.240321683 | 0.2438617 |
| 0.81605436 | 1 | 0.35014108 | 0.32478465 | 0.27789617 |
| 0.24744599 | 0.35014108 | 1 | 0.942558624 | 0.26752607 |
| 0.24032168 | 0.32478465 | 0.94255862 | 1 | 0.10669593 |
| 0.2438617 | 0.27789617 | 0.26752607 | 0.106695931 | 1 |
- It assymes that YOURPATH has a folder name inputHaemRBC and there should exists set of files in YOURPATH/inputHaemRBC/ where each file should be name as follow:
HaemGenRBC_HB.txt HaemGenRBC_HCT.txt HaemGenRBC_MCH.txt
We assume that HB, HCT, and MCH are the name of phenotypes. In other words the name of files in YOURPATH/inputHaemRBC/ should match the header of correlation.data.