CLARK detecting less number of viruses in metavirome dataset

[Rashedul Islam]

To CLARK team,

I am using CLARK to find the taxonomy of viruses in metavirome (SRR10281034) dataset. I assembled the reads into 122,545 contigs (attached). However, I found only 0.5% contigs were predicted to be virus which is surprising to me (output attached). I was expecting at least 10% of the contigs to be assigned to the virus taxonomy. Do you have any comments or suggestions?

The commands used:

#Download
wget -N http://clark.cs.ucr.edu/Download/CLARKV1.2.6.1.tar.gz

#Extract
tar -xzvf CLARKV1.2.6.1.tar.gz

#Install
./CLARKSCV1.2.6.1/install.sh

#Create clark database
./CLARKSCV1.2.6.1/set_targets.sh clarkdb viruses --species

#Calssfiy 

classify_metagenome.sh -O contigs.fa -n 64 -R my_results 

Best

Rashedul