tally_Cogent_contigs_per_family.py error

[jodithea]

I am trialling Cogent on the test_human.fa data as shown on the Wiki, though I did not use the reference genome hg38 for the coding genome reconstruction as I do not have a reference genome for my data and wanted to check I could do everything without it.

python /path/Cogent/helper_scripts/tally_Cogent_contigs_per_family.py
test_human/ hg38 human_output

The above works well. I used hg38 as a placeholder (I don’t have the reference genome hg38 in my wd on purpose), as I received errors when not including this ref genome placeholder. This worked nicely with the human_output.family_summary.txt:

gene_family input_size num_Cogent_contigs num_genome_contig genome_cov genome_acc genome_chimeric genome_contigs test_human/1_0 9 4 0 0.00 0.00 False test_human/0_0 10 1 0 0.00 0.00 False

However, when I include blastn in this step:

python /path/Cogent/helper_scripts/tally_Cogent_contigs_per_family.py
test_human/ hg38 human_output
--blastn in.fa.blastn

I receive the error: Traceback (most recent call last): File "/apps/unit/RavasiU/cogent/8.0.0/Cogent/helper_scripts/tally_Cogent_contigs_per_family.py", line 62, in main(args.cogent_dir, args.genome, args.output_prefix, args.genome2, args.blastn_filename) File "/apps/unit/RavasiU/cogent/8.0.0/Cogent/helper_scripts/tally_Cogent_contigs_per_family.py", line 41, in main sp.tally_for_a_Cogent_dir(d, writer1, writer2, genome1, genome2, blastn_filename) File "/hpcshare/appsunit/RavasiU/cogent/8.0.0/Cogent/helper_scripts/tally_Cogent_results.py", line 191, in tally_for_a_Cogent_dir best_of = read_blastn(os.path.join(dirname, blastn_filename), qlen_dict) File "/hpcshare/appsunit/RavasiU/cogent/8.0.0/Cogent/helper_scripts/tally_Cogent_results.py", line 44, in read_blastn if e < best_of[seqid]: best_of[seqid] = (e, name) TypeError: '<' not supported between instances of 'float' and 'tuple'

And the human_output.family_summary.txt is empty, but does now include the blastn headers:

gene_family input_size num_Cogent_contigs num_genome_contig genome_cov genome_acc genome_chimeric genome_contigs num_blastn blastn_best

I checked, and the in.fa.blastn file is in each of the cogent directories. Do you have any suggestions of what could be going wrong here?

Thank you!