Retention Index Correction

This part is the most important method in BinBase. Basically it makes sure that all the retention times are comparable with each other.

After testing several methods we decided to use 13 internal standards, which are spread out all over the chromatogram and lock the retention times with this. This allows us to compare completely different kind of samples.

The current approach is to filter the chromatogram to get only the spectra which have a certain base peak and ratio between specific masses. Since we can't really use the similarity at this point. This reduces the amount of information by ca 90%.

Once this progress is done we use the similarity and find the highest possible match against the library of retention index standards. And use this hit to compare the relative distance between the library massspec of the best hit and the library massspec of the possible hit. If this is in a defined range we use this as possible retention index marker.

After we make sure that all the markers are found and are found in order we continue and check that a certain number of standards is found. There is also an optional check which specify that certain standards have to be found for a successful correction.

The final step is to apply a polynomial regression to the chromatogram and to correct the retention times in this way.

Now if the algorithm does not find enough standards in the sample it tries to correct the sample with another sample from the same day and if this does not work we look +/- 10 days for a fitting sample.

If this fails too, it uses the curve of the standards it found, to provide at least some corrected data, which are bad as best.