1.Identification of gene families in barley

The amino acid sequences of the newly updated barley genome were acquired from e!DAL(https://doi.org/10.5447/ipk/2019/8).For characterization of integrated and exact candidate mTERF family numbers in barley genome,the file named mTERF.hmm which was downloaded from the Pfam database (http://pfam.xfam.org/) by searching the Pfam Accession Number of mTERF(PF02536).The HMMER 3.0 package was used to perform with Hidden Markov Model (HMM) analysis in terms of default parameters.The candidates were checked using HMMER and Pfam database to confirm the the existence of mTERF domain. The acquired mTERF genes were used as queries in BLASTN similarity search against the EST sequences of barley to verify the existence.Predicted molecular weight(MW), number of amino acids,theoretical pI(isoelectric point) and grand average of hydropathicity(GRAVY) of mTERF proteins were investigated with the online tool ExPASy(https://web.expasy.org/protparam/).Information about subcellular localization of HvmTERF genes were detected using Cell-PLoc 2.0(http://www.csbio.sjtu.edu.cn/bioinf/Cell-PLoc-2/).(Reference: Kuo-Chen Chou, and Hong-Bin Shen, Cell-PLoc 2.0: an improved package of web-servers for predicting subcellular localization of proteins in various organisms, Natural Science, 2010, 2: 1090-1103)The chromosomal localization information was obtained from genome annotation profiles.

2.Sequence alignment and phylogenetic analysis

Amino acid sequence alignments of the full-length HvmTERF proteins were generated using ClustalX v.1.83.Evolutionary analyses were conducted by construct a phylogenetic tree based on the statistical method of neighbor-joining. Bootstrap analysis was carried out with 1000 trials to ensure statistical reliability.

3.Conserved motifs, gene structure and cis-element analysis

The MEME program v5.1.1 (http://meme-suite.org/) with default parameters except the parameter of motif number was arranged as 8,which was used to execute the detailed analysis of conserved motif domains among HvmTERFs.On the basis of barley GFF3 annotation file,exon-intron structure of HvmTERF genes were displayed on the online website Gene Structure Display Server(http://gsds.cbi.pku.edu.cn/).The DNA sequences 1500 bp upstream of initiation codon of each mTERF gene were submitted to the PlantCARE tool (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) for cis-elements investigation.

4.Gene duplication analysis estimation of Ka/Ks ratios

To define gene duplication events,the following criteria were defined:(1)the longer gene was covered more than 80% of the alignment;(2)the identity of genes in the aligned region > 65%;and (3)a block of duplications should contain a minimum of two genes.The duplicated amino-acid sequences were aligned via Clustal Omega v 1.2.4 software.To evalueate the codon evolutionary rate on duplicated genes,the codeml tool, which embedded in PAML interface software was utilized to calculate the rates of Ka/Ks.

5.Collinearity analysis

mTERF protein sequences of Arabidopsis thaliana,Brachypodium distachyon,Vitis vinifera,Oryza sativa、Zea mays,were acquired from the Ensemble Plants database （http://plants.ensembl.org/index.html）by using the method of identification of barley mTERF genes.Homologous mTERF gene pairs of barlay and species were investigated by Perl script and Inparanoid v4.1..Circos v0.67 was employed to reveal the pairs. Meanwhile,the Ka/Ks ratio was extracted via above methods.