pip missing version

HI Liam,

Recently I have had an issue with some query proteins, looking at the output notes for the proteins with an issue I am getting the following message:
Not making unique seq fasta for SpeC_seqs_and_ref_nuc local variable 'i' referenced before assignment

This time I had an issue with 3 out of 23 proteins. I have checked the inputs, which are fine, no odd characters . Also the proteins that have an issue changes across different runs.

Any insights on what is going on and how to fix it?
Please let me know if you need any more information.

Thank you,
Jacqui

Currently have to run run an empty dir due to loose glob.

File "/home/liam/.local/bin/screen_assembly3.py", line 227, in main
    fin = open(any_file, "rt")

Picks up input dirs if any. Make glob more stringent and not pass anything in flags -i and -q

test

Hi,

If using the -i flag for the --percent identity as listed in the wiki it will automatically default to 80 as that flag is already in use for the input folder. Not sure if its an issue in the actual code or not but using --percent_identity instead works.

As per https://stackoverflow.com/questions/16254191/python-rpy2-and-matplotlib-conflict-when-using-multiprocessing plotting in multiple threads doesn't work on Mac. Need to go back to one thread (shame as this is slow). AA variation plot currently only works on linux

Hi, Liking the tool, but have been having a bit of trouble getting the alignment and plotting function working. I'll post separate issues for each thing I have managed to troubleshoot so far.

In the code you call MUSCLE aligner however this isn't listed as a dependency in the wiki. ClustalO is listed and is called following so I'm not sure if MUSCLE has been removed but is still in the code or should be a listed dependency.

in analysis.py lines 766-792

def versions(args):

    '''
    Save wrapped program version to text
    '''

    with open('versions.txt', 'w') as fout:
        fout.write(str(args)+'\n')
        call(['blastn', '-version'], stdout=fout)
        if args.aln:
            call(['muscle', '-version'], stdout=fout)
        #if args.raxml:
        #    call([args.raxml_executable, '-v'], stdout=fout)
        if args.IQtree:
            call(['iqtree', '-version'], stdout=fout)

def clustal(args, query, seq_type):

    if not os.path.exists(query + '_' + seq_type + '_non_redundant.aln'):
        call(['clustalo',
            '-i', query + '_' + seq_type + '_non_redundant.fasta',
            '-o', query + '_' + seq_type + '_non_redundant.aln',
            '--threads', '1'])#seems faster than just giving clustal muli threads

    if args.plots:
        if seq_type == 'aa':
            gap_plot(args, query, seq_type)

Hi, I have previously run the script with nucleotide queries which has been working fine however I tried to run with some protein queries and it ran for some but for others I get the following error message

"Not making unique seq fasta for BCAL1869_seqs_and_ref_nuc local variable 'i' referenced before assignment"

The issue appears to only effect the sequences that are of concern (the csv and box and carriage plots work for the other queries) and when isolated from the multi fasta into a shorter list the same error occurs. I had a quick dig around and suspect that there might be an indentation/ local variable call issue at line 254 of analysis (if i > 1) causing the function responsible for creating the seqs_and_ref_nuc file to throw up the error but not entirely sure.

I am also not sure what it would be that is causing some sequences to fail but others succeed which I assume is a different issue that may be specific to my dataset. In case you want to see if that issue is reproducible here is the link to the sequences I have been using

https://www.burkholderia.com/downloads/burkholderia/bgd_r_9_1/Burkholderia/complete/fna-complete.tar.gz
(I needed to alter file names to remove # and reduce file name length).

fasta file is also attached in .txt format.
mixed_success_fail.txt

def make_fasta_non_redundant(args, query, seq_type):

    '''
    Boil each multi fasta down to unique seqs for every query - at aa level
    '''
    try:
        redundant_map = collections.defaultdict(list)
        for i, record in enumerate(SeqIO.parse(query + '_seqs_and_ref_' + seq_type + '.fasta','fasta')):
            seq_str = str(record.seq)
            redundant_map[seq_str].append(record.id)


        #write non redundant fasta
        fout= open(query + '_' +seq_type+'_unique_seqs.fasta','w')
        if i > 1:
            for i, seq in enumerate(redundant_map):
                if 'ref' in redundant_map.get(seq):
                    fout.write('>0 ' + ','.join(redundant_map.get(seq))+'\n')
                else:
                    fout.write('>'+ str(i + 1) + ' ' + ','.join(redundant_map.get(seq))+'\n')
                fout.write(seq+'\n')
        fout.close()
    except Exception as e:
        print ('Not making unique seq fasta for',query + '_seqs_and_ref_' + seq_type, e)

Hi Shimbalama,
Could you please change your iTOL heatmap file so that the genes with the most hits are shown first, then the others in descending order? Cheers

Hi,

When I try the plot function (on MacOSX) I get the following error

#code run
screen_assembly3.py -t 1 -l 95 --percent_identity 95 -o -k -q genes.mfa -i contigs/

Plotting...
Traceback (most recent call last):
File "/Users/user/.local/bin/screen_assembly3.py", line 232, in
main()
File "/Users/user/.local/bin/screen_assembly3.py", line 205, in main
pool.map(iplot.plot_vars, tmp)
NameError: name 'iplot' is not defined

looking through the script I couldn't see where the module iplot (or plot which is on line 211) is defined (I'm assuming they are calling a module from matplotlib but wasn't sure). Not sure if there is a library that I haven't loaded but it looks like there needs to be a statement importing something as iplot at the beginning of the script.

        if args.plots:
            print ('Plotting...')
            if args.operon:
                with Pool(processes=int(args.threads)) as pool:
                    tmp = [(args, query, 'nuc') for query in query_seqs]
                    pool.map(iplot.plot_vars, tmp)
            else:
                variant_types(args, assemblies) 
                with Pool(processes=int(args.threads)) as pool:
                    tmp = [(args, query, 'aa') for query in query_seqs]
                    pool.map(plot.plot_vars, tmp)   
            var_pos_csv(args, ana.blast_type)

If it helps to troubleshoot it appears some of the plots are working as I get the box_and_carriage plots.svg outputs but I don't get any of the snp/ gap variation plots. Not a major concern but