grexor / pybio Goto Github PK

Move bin parts to a new format: pybio action

pybio config . doesn't work

For Ensembl we have:

pybio ensembl homo_sapiens 109

For a custom genome, we could have:

pybio add_genome -species homo_sapiens -version custom1 -gtf path_to_gtf_file -fasta path_to_fasta_file

Remove pybio config user input prompt, problematic for Dockerization

Hi,

The genomes function find_linear throws the following error:

.../genomes/init.py", line 475, in find_linear
index = bisect.bisect_left(genome_linear, (pos,"",""))
TypeError: '<' not supported between instances of 'int' and 'str'

A workaround that worked for me is to use tuples of two integers:

genome_linear_coords = [i[0:2] for i in genome_linear] 
index = bisect.bisect_left(genome_linear_coords, (pos,pos))

Best,
Daniel

Create a virtual GTF file with a few genes, both strands, one chromosome. Iterate over all the chromosome and for each position determine if pybio.genomes.annotate works as expected (produce truth array and compare).

• convert exons set to list and transcript set to list
• sort exons in transcripts based on strand

Hi Gregor,

I have been using your tool to develop a simplified version of the tool for my requirement of figuring out differentially expressed polyA sites. Taking aid of your recent documentation, I have been able to generate the polyA atlas. Currently I am figuring out the call annotation(poly_id) in the process function of apa/polya directory.

What I understand is:

annotate(poly_id) calls pybio.genomes.annotate,
pybio.genomes.annotate calls load(species) which requires the genes, intervals and linear json.

genes, intervals and linear json are created in prepare and prepare_gtf.
Please let me know if I am going the right way? If yes, which function should I use: prepare or prepare_gtf?

For software that does not handle gz files, keep uncompressed gtf+gff annotations.

http://plants.ensembl.org/Arabidopsis_thaliana/Info/Index

If several Ensembl genome versions for a species are downloaded and prepared, always work with the latest by default (sort by version desc).