Utilization of the Proteome Data Deposited in SRMAtlas for Validating the Existence of the Human Missing Proteins in GPM

The
Human Proteome Project (HPP) has made great efforts to clarify
the existing evidence of human proteins since 2012. However, according
to the recent release of neXtProt (2019–1), approximately 10%
of all human genes still have inadequate or no experimental evidence
of their translation at the protein level. They were categorized as
missing proteins (PE2–PE4). To further the goal of HPP, we
developed a two-step bioinformatic strategy addressing the utilization
of the SRMAtlas synthetic peptides corresponding to the missing proteins
as an exclusive reference in order to explore their natural counterparts
within GPM. In the first step, we searched the GPM for the non-nested
SRMAtlas peptides corresponding to the missing proteins, taking under
consideration only those detected via ≥2 non-nested unitypic/proteotypic
peptides “Stranded peptides” with length ≥9 amino
acids in the same proteomic study. As a result, 51 missing proteins
were newly detected in 35 different proteomic studies. In the second
step, we validated these newly detected missing proteins based on
matching the spectra of their synthetic and natural peptides in SRMAtlas
and GPM, respectively. The results showed that 23 of the missing proteins
with ≥2 non-nested peptides were validated by careful spectral
matching.