Usage of ballcools query

[xixueyu96]

Hi @jksr,

First of all, thanks for developing ballcools and I've been actively testing it in my pipeline. I have some doubts about the usage of ballcools query.

$ ballcools --version
0.9.9
$ ballcools query --help
Query info from a BAllC file
Usage: ballcools query [OPTIONS] ballcpath genomeranges...

Positionals:
  ballcpath TEXT REQUIRED     BAllC file path
  genomeranges TEXT ... REQUIRED
                              Genome range of interests. Supported formats are chrX, chrX:    chrX:XXX-XXX, chrX:-XXX, chrX:XXX-. Multiple ranges can be separated by space.

Options:
  -h,--help                   Print this help message and exit
  -c,--cmetapath TEXT         Output path
  -x,--c_context TEXT [*]     Filter Cs only in desired context. The context code is three letters from IUPAC nucleotide symbols. The first letter should always be C.For example, the c_context CGN matches CGA, CGT, CGC, CGG, and CGN, and only C records in these contexts will be displayed.
  -s,--c_strand TEXT [.]      Filter Cs strandness. Use "+" for only Cs on the plus/Watson strand, "-" for only Cs on the minus/Crick strand, "." or "*" for both strands.
  -o,--operation TEXT [query] 
                              Perform a selected operation on the querying genome range. Currently supported operations are:
                              "query" to display the records in the specified range (ie. no operation will be performed). The output format is of [chrom pos strand context mc cov 1];
                              "frac" to calculate the mean of mc and cov values in the specified range. The output format is of [#C mCFrac];
                              "sum" to calculate the sum of mc and cov values in the specified range. The output format is of [#C sum_mc sum_cov];

Q1: what if I want to filter Cs for globally? since the genomeranges is required, is there a more easy way to declare the whole genome range?
Q2: is the -c,--cmetapath meant to be the -o, --outputpath? It's a bit confusing for me about the meaning of this parameter.

Thx,

Sherry

[jksr]

@xixueyu96
hi Sherry,
sorry for the very late reply, I missed the reminder email of this issue since I was in parental leave.

Q1: you could use ballcools query file.ballc "*" to list all the records in the file. you can add additional filters as you like. the quote markers are necessary. otherwise, most shell will expand the wildcard * as the all the filenames in the current folder.

Q2: sorry, it was a typo. and it has been corrected in the latest version. btw, there are some convenient functions update in the latest version as well

Thank you for using ballcools. I'm more than happy to talk if there is anything I can help to streamline ballcools in your pipeline. Hope this is not too late:)

[xixueyu96]

@jksr
hi jksr,
sorry for late reply. It seems that by ballcools has been the latest version according to the README. I reinstall the ballcools by conda recently, and still got the same version and the same help message. So I still have two questions as follows.

Q1: How to install the lastest version of ballcools?
Q2: It seems that the --operation argument does not output the expected result, could you please provide some practices or tutorials?

thank you again for developing ballcools, and congrats for your publication!

$ ballcools query -o "query" test_cell.genome_CpG.ballc chrX:52988136-53021659
chrX    53015002        2       2
chrX    53015005        2       2
chrX    53018309        1       1
chrX    53018400        0       1
chrX    53019225        2       2

$ ballcools query -o "frac" test_cell.genome_CpG.ballc chrX:52988136-53021659
chrX    53015002        2       2
chrX    53015005        2       2
chrX    53018309        1       1
chrX    53018400        0       1
chrX    53019225        2       2

$ ballcools query -o "sum" test_cell.genome_CpG.ballc chrX:52988136-53021659
chrX    53015002        2       2
chrX    53015005        2       2
chrX    53018309        1       1
chrX    53018400        0       1
chrX    53019225        2       2

[yoeHuuui]

@jksr
Hi! Really nice tools—thanks for sharing them.
When testing ballcools query, we found that the operation argument does not seem to work. Also, the README mentions that ballcools query supports the -b option for BED file input, but we could not find this parameter in version 0.9.9.

Here is the test log:

$ ballcools query --operation query ballc/test.ballc chr1:1-1000000 | head -n5
chr1    102988  0       4
chr1    102995  0       4
chr1    135028  4       4
chr1    135031  4       4
chr1    135202  4       4

$ ballcools query --operation frac ballc/test.ballc chr1:1-1000000 | head -n5
chr1    102988  0       4
chr1    102995  0       4
chr1    135028  4       4
chr1    135031  4       4
chr1    135202  4       4

$ ballcools query --operation sum ballc/test.ballc chr1:1-1000000 | head -n5
chr1    102988  0       4
chr1    102995  0       4
chr1    135028  4       4
chr1    135031  4       4
chr1    135202  4       4

$ ballcools query --operation sum ballc/DRX015008.ballc -b ../sequences.1K.bed | head -n5
The following argument was not expected: -b
Run with --help for more information.

$ ballcools -v
0.9.9

[yoeHuuui]

I have fiexed 'operation ' problem with -c parameter, but I also failed to use the -b parameter.