It is not uncommon that bioinformatics programs produce highly verbose and unstructured output that is very time consuming to read manually. Quite often, only a small subset of the results are needed to evaluate the evidence.
- In this exercise you will work with a huge results archive that you can download from here (1st half) and here (2nd half). The data has been produced from 500 bootstrap replications and each replicate directory contains a single file of interest: psn.lst. From this file you should extract and report the following: i) Minimization: whether it was successful or not; ii) OFV-value (called OBJV in the file) and the parameter estimates for Thetas and Etas. The aim is to produce a single table in the form of a CSV-file where the columns are:
Perl1Replicate Successful OBJV TH1 TH2 TH3 TH4 TH5 TH6
The pairwise Eta matrix can be extracted and dumped into a separate file, one for each bootstrap replicate.
- Linearize the pairwise Eta matrix so that you print additional columns in the main CSV-file, one column per Eta pair (ETA1vsETA2, ETA1vsETA3 and so on).