Software installation instructions

[tavareshugo]

@lkalmar can you list here on this issue all the software that is used in the course?
So we can keep track of things for future iterations of the course.

[lkalmar]

R packages:
dada2, phyloseq, Biostrings, ggplot2, reshape2, readxl, tydiverse

Command line applications:
fastqc, multiqc, cutadapt, trimmomatic, bowtie2, samtools, metaphlan, mash, SPAdes, clumpify.sh (part of the bbmap package), flash (for merging reads), maxbin2, checkm (database!), gtdbtk (database!), prokka, abricate (database!)

[tavareshugo]

Can I check:

1. It's checkm (version 1), not checkm2?
2. Do you have commands/links to download all those databases?

[lkalmar]

Can I check:

1. It's checkm (version 1), not checkm2?
2. Do you have commands/links to download all those databases?

1. it is the original checkm, that is still the gold standard, but I will look into checkm2
2. I think all of these are conda installable, but to be sure, I will collect all the installations here.

[tavareshugo]

For point 2. I just meant the databases, not the software itself.
For example, with CheckM2 they have a command checkm2 database --download --path <output>.

I don't think CheckM (version 1) has a command, but there is this: https://data.ace.uq.edu.au/public/CheckM_databases/
Is that the correct database to download?

For the other programs, I don't know if the databases come with the software or if they need to be installed separately.

[tavareshugo]

gtdbtk database can be obtained with the download-db.sh command

[tavareshugo]

abricate just seems to have the databases as part of the installation.

@lkalmar I've updated the data and setup page, would you mind revising before I close this issue?

[lkalmar]

Yes, checkm has a database and you need to set the database path. Abricate comes with the database, but there is a way to update those here

[lkalmar]

Otherwise, the update is perfect, please close the issue

[tavareshugo]

Reopening as we are missing MetaPhlan database.

In this page they recommend for conda installations:

metaphlan --install --bowtie2db <database folder>

Does this look right @lkalmar?

In the future we would then need to adjust the materials to point to a --bowtie2db folder that we decide to save the database into.

[lkalmar]

You either install the DB in your conda / miniconda / mamba / micromamba folder with the simple command metaphlan --install (not recommended on the HPC, but on own computer or in-house server this is simpler).

Or, you define the database path with the above mentioned metaphlan --install --bowtie2db <database folder> but in that case you have to define the database path during the run.

[tavareshugo]

Installation scripts from metaphaln and gtdb-tk are not reliable.

• MetaPhlan: follow instructions here to find the FTP link to their databases. Will need to dig into their install scripts to figure out exactly what files we need.
• GTDB-Tk: Follow instructions to download manually

[tavareshugo]

it's a bad idea to have everything in the same environment, due to dependency conflicts (e.g. an old version of maxbin).

Update the instructions to have each software in a separate environment.

[lkalmar]

Maybe not each, than we end up with a huge number of envs...