sinamajidian / fastoma_tree

FastOMA is a fast pipeline for species tree inference using placement in Hierarchical Orthologous Groups (HOGs).

FastOMA (under development)

FastOMA is a novel and fast pipeline for species tree inference using placement in Hierarchical Orthologous Groups (HOGs).

prerequisites

Please install these

• OMAmer using pip install omamer.
• Pyoma using pip install pyoma
• [zoo wrapper] ??
• mafft

and packages datetime, numpy, ast, pickle .

Download these

• OMAmer dataset (10Gb) ??
• [HOH-OG-map.dic] >> my zenodo (100Mb)
• OMA dataset and its index file (95+60Gb).

cd test_data; mkdir dataset; cd dataset
wget https://zenodo.org/record/4593702/files/LUCA.h5     ??
wget https://omabrowser.org/All/OmaServer.h5
wget https://omabrowser.org/All/OmaServer.h5.idx

Be careful about different OMA releases. The OMAmer dataset and OMA dataset should be from the same release since HOG id changes in every release.

Pipeline

Consider you have the proteome of two species as two fasta files WoodlandKingfisher.fa (Halcyon senegalensis, HALSEN) and AtlanticCanary.fa (Serinus canaria, SERCAN) which are provided in the test_data folder. Note that the fasta file shouldn't contain U, you could replace it with X with sed s/U/X/g otherwise.

The first step is to find the placement of each protein of each species in the HOG dataset. To do so , we use OMAmer. We need to download the OMAmer dataset. Then, we can run OMAmer:

cd ..; mkdir hogmap

number_threads=20

omamer search --db  dataset/LUCA.h5 --query proteome/WoodlandKingfisher.fa --nthreads ${number_threads}  --out hogmap/WoodlandKingfisher.hogmap
omamer search --db  dataset/LUCA.h5 --query proteome/AtlanticCanary.fa --nthreads ${number_threads}  --out hogmap/AtlanticCanary.hogmap

After few minutes, the HOG maps of species are ready in the hogmap folder.

number_threads=20
python fastoma-tree.py  ./test_data/dataset/   ./test_data/  ${number_threads}

The second argument should contain two folders proteome and hogmap. The output will be the multiple sequence alignment supermatrix.

Finally, you could run a tree inference method

fasttree _20_msa_concatanated.txt

(CHLSB:0.007967770,PAPAN:0.009104705,(HUMAN:0.012101579,((AtlanticCanary:0.327152404,(MOUSE:0.121086106,CHILA:0.087558595)0.971:0.015068557)0.167:0.000847169,WoodlandKingfisher:0.353362694)1.000:0.041260314)1.000:0.010528717);

bird data set

We used following thresholds for bird dataset:

keep_og_treshold_species_query =   360
threshold_least_query_sepecies_in_OG = 320
kept_oma_species_num = 20
ogs_keep_number = 100

done!