What I Learned from Analyzing Accurate Mass Data of 3000 Supporting Information Files

A Python script for the systematic, high-throughput analysis of accurate mass data was developed and tested on more than 3000 Supporting Information (SI) PDFs from Organic Letters. For each SI file, quadruplets of molecular formula, measured ion, e.g., [M + Na]+, and reported calculated and found masses were extracted and analyzed. Interestingly, only 40% of the files containing readable accurate mass data were both internally consistent and in compliance with The ACS Guide to Scholarly Communication. The analysis revealed unexpected errors and provided actionable advice on how to improve data quality.