Benchmarking the composition-based sequencing approach on high-resolution Orbitrap-HCD data

摘要

The software for automatic de novo sequencing utilizing the composition-based sequencing approach was developed. The input data (the plain text representation of deconvoluted and decharged spectra) might be generated from any raw MS data relatively fast. The performance on the mixed fragmentation spectra set (MH+ 700-1400 Da) is about 2 spectra/minute. Predictive possibility of the software is more than 20%; the main obstacle is the high level of noise peaks or peaks not included in used fragmentation model (eg. internal ions). Analysis of HCD data and simulated CID data give similar results. The predictive performance is slightly higher for simulated CID data, but the quality of result is lower (lower score, more sequence gaps etc.).