*2.7. Bioinformatics*

#### 2.7.1. Data Source

In this experiment, a total of 20 LC–MS/MS runs (81 min runs) were conducted on the Thermo Q-Exactive mass spectrometer. Three extra LC–MS/MS runs represented pools of all samples. A total of 30,000 MS/MS spectra were obtained for each plant species. Protein sequence databases downloaded from the Phytozome protein database on 11 October 2017, held 88,760 proteins for maize and 47,121 proteins for sorghum.

#### 2.7.2. Peptide and Protein Identification Pipeline

Raw data files (spectra) were converted into mzML using the ProteoWizard 3.0 msConvert tool [29]. Peak-Picking and Zlib compression were employed. Database searching employed the MS-GF+ search engine [30] to identify the potential peptides shaped by semi-tryptic specificity, and a 20 ppm precursor tolerance was applied. The data retrieval results were refined by IDPicker (version 3.1) [31] to produce a 2% peptide–spectrum match (PSM) false discovery rate (FDR), with two unique peptides required for each protein.

NCBI BLAST 2.5.0+ makeblastdb [32] was used to index the FASTA sequence databases for ortholog identification between the two species. The blastp program was used to query each sorghum sequence in the maize database and to query each maize sequence in the sorghum database. The generated ortholog files were read in R statistics script and a minimum bit score of 50 was applied. Matches that exceed this threshold were considered true. In cases where multiple matches were found, only the hit with the highest bit score value was retained. Ortholog data and spectral counting tables from IDPicker were read in a script in the R statistical environment to align the spectral count row for each sorghum protein with the spectral count row for the orthologous maize protein. When maize and sorghum orthologs were split into different rows (for example, in the case of paralogs for one species but not the other), the two rows were merged to form one joint row. Proteins lacking orthologs or those with unidentified orthologs were not subjected to further analyses.
