Analysis and supplementary data for Srivastava

1. Download data

Download proteome and annotation (gtf/gff) for each species in the Supplementary file genomes, also found in the input folder.

2. Select one protein per gene

Run 01_longest_prot.R

The output is a fasta file per species including the longest amino acid sequence per gene, rename by gene name. 1gene == 1protein

3. Ortholog analysis

We used OrthoFinder to identify ortholog groups. The concatenated multiple sequence aligment (MSA) of genes in orthogroups with only one copy and present in all species was used to infered the tree with MAFFT v7.453 MBE 30:772-780 (2013); NAR 30:3059-3066 (2002)

4. Identify signal peptides with SignalP 5.0

Nat Biotechnol 37:420-423 (2019)

Use the following command (bash) to run SignalP-5.0 locally per genome.

for i in `ls `; do signalp5 -fasta $i -format short > ../signalP_output/${i}_summary.signalp5 ; done

5. Select proteins with signal peptides identified by SignalP-5.0

Run 03_filter_signalp.R

The output is a fasta file per species of proteins with signal peptide.

6. Identify transmembrane topology and signal peptides with Phobius ver 1.01

J Mol Bio 338(5):1027-1036 (2004)

Use the following command (bash) to run phobius-1.01 locally per file of signalp identified proteins.

for i in `ls ../signalP_output/*faa`; do phobius -short ${i} > ${i}_phobius.out ; done

7. Filter out proteins with signalp and transmembrane domains

Run 04_filter_phobius.R

The output is a fasta file per species of proteins with signal peptide but without transmembrane domains.

8. Identify effectors in extracellular fungal proteins with EffectorP v3.0

Mol Plant Microbe Interact 35:146-56 (2022)

Use the following command (bash) to run EffectorP-3.0 locally per file

python3 /mnt/e/software/EffectorP-3.0-main/EffectorP.py -f -i ../phobius/*_signalp.faa_phobius.faa > *.aa_signalp.faa_phobius.faa_fungal_effectorP.out

9a. Obtain CAZyme protein list from JGI

Per JGI genomes, obtain the list of CAZymes and save it in an file, 1 file/sheet == 1 species.

Run 02_jgigff_cazyme.R to make sure all CAZy protein ID are in gff file.

9b. Predict CAZymes with dbCAN

NAR 51:W115-W121 (2023)

Predict CAZyme proteins with dbCAN3 for NCBI proteomes and those species from JGI without a list of CAZymes.

10. Calculate number of proteins per species, 1gene == 1 protein

for i in `ls speciesSelected/*fa*`; do grep -iHEc '>' $i | sed 's/:/\t/g'; done > proteome_size.txt

11. Calculate number of proteins with signal protein and without transmembrane domain, aka the secretome

for i in `ls tmhmm_phobius/*phobius.faa*`; do grep -iHEc '>' $i | sed 's/:/\t/g'; done > secretome_nosignal_nophobius.txt

12. Make the summary graphic

Run 05_graph.R