Error while installing qiime2 with conda in Rstudio server (linux)

[Lotte (Lotte Geerlings)]

Dear all,

I am trying to create a pipeline for my oral health samples to check the Taxonomical classification with Qiime. My university has provided a server (Rstudio) in which I use the terminal for Linux & conda:
Current conda install:

         platform : linux-64
    conda version : 3.7.0

conda-build version : not installed
python version : 3.4.1.final.0
requests version : 2.4.1
root environment : /opt/miniconda (read only)
default environment : /home/lotte_geerlings/envs/taxonomy
envs directories : /home/lotte_geerlings/envs
/opt/miniconda/envs
package cache : /home/lotte_geerlings/envs/.pkgs
/opt/miniconda/pkgs
channel URLs : https://conda.binstar.org/biocore/linux-64/
https://conda.binstar.org/qiime2/linux-64/
https://conda.binstar.org/qiime2/label/r2018.11/linux-64/
https://conda.binstar.org/conda-forge/linux-64/
https://conda.binstar.org/bioconda/linux-64/
http://repo.continuum.io/pkgs/free/linux-64/
http://repo.continuum.io/pkgs/pro/linux-64/
config file : /home/lotte_geerlings/.condarc
is foreign system : False

– Now that I am creating the pipeline however, I get stuck with Qiime. I keep getting error
messages.

QIIME2

SUDO APT-GET INSTALL NCBI-BLAST+
CONDA CREATE -N QIIME2-2017.10 --FILE
HTTPS://DATA.QIIME2.ORG/DISTRO/CORE/QIIME2-2017.10-CONDA-LINUX-64.TXT

For which I get this error message:

Error: No packages found in current linux-64 channels matching: @explicit
Did you mean one of these?
implicit
You can search for this package on Binstar with
binstar search -t conda @explicit
You may need to install the Binstar command lineclient with

I have also tried doing this by making a bash script (.sh) and then as
follows:

#!/bin/bash

chmod +x qiime2_install.sh

name: qiime2-2018.11

channels:

conda config --add channels qiime2/label/r2018.11
conda config --add channels qiime2
conda config --add channels conda-forge

(etc.) But I keep running into this error:

Fetching package metadata: …
Solving package specifications: .

All requested packages already installed.

packages in environment at /home/lotte_geerlings/envs/taxonomy:

arb-bio-tools 6.0.6
h5901010_5
Fetching package metadata: …
Solving package specifications:
Error: Could not find some dependencies for bioconductor-biobase
2.38|2.38.0*: r 3.2.2*, r 3.3.1*, _r-mutex 1.* anacondar_1, jpeg >=9c,<10a
You can search for this package on Binstar with
binstar search -t conda r 3.2.2*
You can search for this package on Binstar with
binstar search -t conda r 3.3.1*
You can search for this package on Binstar with
binstar search -t conda _r-mutex 1.* anacondar_1
You can search for this package on Binstar with
binstar search -t conda jpeg >=9c,<10a
You may need to install the Binstar command lineclient with
conda install binstar

I did already install binstar and jpeg version, but right now I feel like
doing a bash version by hand would be far too much work and that there
should be a more simple solution to this problem…

Can anyone help me with this?

Thanks in advance :-)!

[jwdebelius (Justine Debelius)]

Hi @Lotte,

A couple questions.

First, do you have another way to access your server? I’m concerned that accessing through R studio might do something to the delicate ecosystem of R in conda, particularly with regard to the way R packages get sourced from conda channels.
I know a lot of people who use Windows use PuTTy which is a free client.

Second, it looks like your command line prompts and your bash script are trying to install two different enviroments. Do you want the November 2018 release, or the October 2017?

Finally, could you please run the command,
conda list -e

and send the output? It looks like you may have things already installed, but that will give us the list and versions.

Best,
Justine

[Lotte (Lotte Geerlings)]

Hi @jwdebelius!

I just started coding so thanks for the help! It seemed that the problem was indeed the combination of Rstudio and conda, once I got a new server that only contains Ubuntu and conda I managed to create a working variant of Qiime.

As for different environments, that was rather sloppy of me. It’s a work in progress.

Thanks again Justine!

Best,

Lotte