count.seqs(shared=...) does not transpose a shared file but returns a count_table

[Shaunson26]

count.seqs(shared=...) does not transpose a shared file but returns a count_table

fake.shared

label  Group   numOtus Otu01   Otu02   Otu03   Otu04   Otu05   Otu06   Otu07   Otu08   Otu09   Otu10
0.10   forest  10  0   5   2   3   1   1   3   3   1   0
0.10   pasture 10  7   2   5   1   3   2   0   0   1   2

mothur

count.seqs(shared=fake.shared)
0.10

Output File Names:
fake.0.10.count_table

cat the result

cat fake.0.10.count_table
#Compressed Format: groupIndex,abundance. For example 1,6 would mean the read has an abundance of 6 for group forest.
#1,forest       2,pasture
Representative_Sequence total   forest  pasture
Otu01   7       2,7
Otu02   7       1,5     2,2
Otu03   7       1,2     2,5
Otu04   4       1,3     2,1
Otu05   4       1,1     2,3
Otu06   3       1,1     2,2
Otu07   3       1,3
Otu08   3       1,3
Otu09   2       1,1     2,1
Otu10   2       2,2

[mothur-westcott]

I think you are looking for the tidy shared format, https://mothur.org/wiki/make.shared/#converting-a-shared-file-to-a-tidy-shared-file.

[Shaunson26]

Hi Sarah,

I had remembered the transpose method from years ago, and indeed wanted it for faster import into R. I since discovered the tidy-shared file format and will use that from here on - pivoting in R take only seconds.

I guess now the issue is whether the section https://mothur.org/wiki/count.seqs/#shared should be removed or the function bug addressed? (I prefer the former!)

Shaun

[mothur-westcott]

Thanks for bringing this to my attention. The count.seqs command has a minor bug. The compress parameter is not working for the shared option. The compress option allow you to print the count table in full format or sparse format. The default is sparse, which is what you are seeing above.

mothur > count.seqs(shared=fake.shared)

should produce:

#Compressed Format: groupIndex,abundance. For example 1,6 would mean the read has an abundance of 6 for group forest.
#1,forest 2,pasture
Representative_Sequence total forest pasture
Otu01 7 2,7
Otu02 7 1,5 2,2
Otu03 7 1,2 2,5
Otu04 4 1,3 2,1
Otu05 4 1,1 2,3
Otu06 3 1,1 2,2
Otu07 3 1,3
Otu08 3 1,3
Otu09 2 1,1 2,1
Otu10 2 2,2

mothur > count.seqs(shared=fake.shared, compress=f)

should produce:

Representative_Sequence total forest pasture
Otu01 7 0 7
Otu02 7 5 2
Otu03 7 2 5
Otu04 4 3 1
Otu05 4 1 3
Otu06 3 1 2
Otu07 3 3 0
Otu08 3 3 0
Otu09 2 1 1
Otu10 2 0 2

As a workaround you can run the command twice:

mothur > count.seqs(shared=fake.shared)
mothur > count.seqs(count=current, compress=f)