interproscan.sh: allow user-specified database path

[nick-youngblut]

It appears that the database paths are hardcoded in interproscan.sh. For example:

export EMBOSS_ACDROOT=bin/nucleotide
export EMBOSS_DATA=bin/nucleotide

This forces a non-standard conda install. For example:

mamba create -n my_conda_env bioconda::interproscan

...produces the following warning:

######################################
# First time usage please README !!! #
######################################

The databases are huge and consequently not shiped within this installation.
Please download and install the Databases manually by following the commands below:
!!! /!\ Edit the 2 first lines to match the wished version of the DB /!\ !!!

Commands:
=========
# See here for latest db available: https://github.com/ebi-pf-team/interproscan or http://ftp.ebi.ac.uk/pub/software/unix/iprscan/5/
# Set version
version_major=5.54
version_minor=87.0

# get the md5 of the databases
wget http://ftp.ebi.ac.uk/pub/software/unix/iprscan/5/-/interproscan---64-bit.tar.gz.md5
# get the databases (with core because much faster to download)
wget http://ftp.ebi.ac.uk/pub/software/unix/iprscan/5/-/interproscan---64-bit.tar.gz
# checksum
md5sum -c interproscan---64-bit.tar.gz.md5
# untar gz
tar xvzf interproscan---64-bit.tar.gz
# remove old DB
rm -rf /tmp/global2/nyoungblut/code/dev/ll_pipelines/llcds/tmp/interproscan_env/share/InterProScan/data/
# copy past the new db
cp -r interproscan--/data /tmp/global2/nyoungblut/code/dev/ll_pipelines/llcds/tmp/interproscan_env/share/InterProScan/


INFO:
====
Phobius (licensed software), SignalP, SMART (licensed components) and TMHMM use
licensed code and data provided by third parties. If you wish to run these
analyses it will be necessary for you to obtain a licence from the vendor and
configure your local InterProScan installation to use them.
(see more information in /tmp/global2/nyoungblut/code/dev/ll_pipelines/llcds/tmp/interproscan_env/share/InterProScan/data/<db>)

It would be helpful if the user can specify the database path(s) versus being forced to find the interproscan install path and move/copy/symlink the data to that location.

[splaisan]

would be great indeed to prevent duplication of the databases, as well as a conda package for the latest version 5.57-90.0 which seems to be so much better (if I knew how to make one I would volunteer but it is not the case)