Closed cmsand closed 4 years ago
This is likely linked to how the correlation is being computed in mzSample->correlation()
.
Everything might be techinically correct, potentially slightly confusing UI.
Note related issue #102 (also has to do with correlation computation)
I can resolve the issue for Valine by decreasing the isotope-parent correlation to -0.1, but that seems to introduce lots of noise into other amino acids. It's an OK work-around for now, but would be good to fix the issue directly.
New approach implemented as of 5a011b4e3fc8289036d74b85fcc4ea6d65b56f6f
Consider the molarity of the adduct when considering formulas - eg [2M+H]+
should enumerate 2*molecular_formula
worth of atoms
Implemented molarity of adduct consideration in maven_core
e1f3b27c82f29754ee8c8ef2e637615aa318b25d
Implemented with bug fixes as of d1e4d095d307e423008ba4a541bda11d7d4753a5
Maven reporting zero values for C12 parent isomer of L-Valine. Spectra view looks OK, though. Manual integration did not fix problem.
Raw data to reproduce bug in Metabolomics Core -> GitHub Issues