Separate output #43
Separate output #43
Conversation
…usSTR into readme_update
|
In addition to separating files, this PR changes how lusSTR deals with missing data. Previously, it would drop any allele with 0 reads. However, this data is necessary for EuroForMix... therefore now lusSTR does not drop any allele from the output. |
|
This is ready for review @standage. I also plan to release the next version of lusSTR... it's been a minute since I last did that. :) |
| data = uas_load(infile, snp_type_arg) | ||
| data_filt = data.loc[data['Reads'] != 0].reset_index(drop=True) | ||
| data_filt = uas_load(infile, snp_type_arg).reset_index(drop=True) |
There was a problem hiding this comment.
Ok, this is where you're retaining alleles with 0 reads.
| if data_filt.loc[j, 'Typed Allele?'] == 'No': | ||
| flag = 'Contains untyped allele' |
There was a problem hiding this comment.
Does the Typed Allele? column refer to whether there were any reads for that allele?
In any case, space and punctuation in column names can be problematic. If you have just added the column in this PR, I'd recommend using IsTyped instead, and boolean values (True/False) rather than "Yes"/"No" strings.
There was a problem hiding this comment.
Or AlleleIsTyped or something.
There was a problem hiding this comment.
The Typed Allele? column is from the Sample Details Report... it doesn't necessarily indicate an allele with 0 reads, but an allele with reads below the various thresholds, so can have a low number of reads as well as 0 (i.e. is the allele considered to be a real allele). The Yes/No is read directly from the Sample Details Report, so I'd prefer to leave that as is, but I can change the column name.
There was a problem hiding this comment.
The
Typed Allele?column is from the Sample Details Report
I see. Maybe worth just leaving it in then...
lusSTR/snps.py
Outdated
| try: | ||
| os.mkdir(output_dir) | ||
| except FileExistsError: | ||
| pass |
There was a problem hiding this comment.
I'd suggest:
os.makedirs(output_dir, exist_ok=True)
This PR provides the option to separate the final output files by sample and output separate files. This is useful when inputting directly into LLAMAS.