Install swarm

[Laura]

Hello everyone!

I'm trying to use the command pick_otus.py with -m swarm, but it's not between the available options. I'm working with QIIME 1.8.0.

I've tried to use the command swarm -h, but it says "No command 'swarm' found", so I guess I have to install it or implement it in qiime. Can you help me with this?

Another question: I'm comparing different otu picking methods (pick_otus.py). Is it better to use a method which creates "seeds" (such as uclust) or the ones which compare with a database? As of now, I would like to compare usearch, trie and cdhit.

Thank you in advance

Laura

[Tomasz]

Hi Laura,

Use `print_qiime_config.py` to check your QIIME components. If possible, I would advise you update to QIIME 1.9+.

As for the clustering question, it really depends on the question you're asking. There's no universally best clustering/otu picking method. If you use closed reference, computations are quick and comparable between datasets, but you're constrained by the contents of your database. Usually, closed reference is fine for fecal samples, or other well-characterized environments. De novo on the other hand, enables you to analyze even a completely unannotated environment, but it's more difficult to compare between different studies, depending on your dataset, it could be associated with a significant computational cost and you don't have taxonomies. You can read more on this in, e.g. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4517945/ (section 5.1.2).

I also passed your question to my colleague, who should be able to advise you on your clustering question better.

t.

[Laura]

This is what I get whith print_qiime_config.py

qiime@qiime-VirtualBox:~$ print_qiime_config.py 

System information

==================

         Platform: linux2

   Python version: 2.7.3 (default, Dec 19 2013, 03:13:59)  [GCC 4.6.3]

Python executable: /home/qiime/qiime_software/python-2.7.3-release/bin/python

Dependency versions

===================

                     PyCogent version: 1.5.3

                        NumPy version: 1.7.1

                   matplotlib version: 1.3.1

                  biom-format version: 1.3.1

                         qcli version: 0.1.0

                QIIME library version: 1.8.0

                 QIIME script version: 1.8.0

        PyNAST version (if installed): 1.2.2

                      Emperor version: 0.9.3

RDP Classifier version (if installed): rdp_classifier-2.2.jar

          Java version (if installed): 1.6.0_27

QIIME config values

===================

                     blastmat_dir: /home/qiime/qiime_software/blast-2.2.22-release/data

                         sc_queue: all.q

      topiaryexplorer_project_dir: None

     pynast_template_alignment_fp: /home/qiime/qiime_software/core_set_aligned.fasta.imputed

                  cluster_jobs_fp: /home/qiime/qiime_software/qiime-1.8.0-release/bin/start_parallel_jobs.py

pynast_template_alignment_blastdb: None

assign_taxonomy_reference_seqs_fp: /home/qiime/qiime_software/gg_otus-13_8-release/rep_set/97_otus.fasta

                     torque_queue: friendlyq

   template_alignment_lanemask_fp: /home/qiime/qiime_software/lanemask_in_1s_and_0s

                    jobs_to_start: 1

                cloud_environment: False

                qiime_scripts_dir: /home/qiime/qiime_software/qiime-1.8.0-release/bin

            denoiser_min_per_core: 50

                      working_dir: /tmp/

                    python_exe_fp: /home/qiime/qiime_software/python-2.7.3-release/bin/python

                         temp_dir: /tmp/

                      blastall_fp: /home/qiime/qiime_software/blast-2.2.22-release/bin/blastall

                 seconds_to_sleep: 60

assign_taxonomy_id_to_taxonomy_fp: /home/qiime/qiime_software/gg_otus-13_8-release/taxonomy/97_otu_taxonomy.txt

[Tomasz]

Hi Laura,

If you had swarm installed along with QIIME it should have been listed in Dependency versions section. I'm not entirely familiar with QIIME legacy versions and I'd advise you update to the latest QIIME1 version, if possible. How to do it is described here: http://qiime.org/install/upgrade.html

t. 

[Jose Antonio Navas Molina]

Hi Laura,

Swarm was added to QIIME in version 1.9.0, so you will need to update QIIME as Tomasz pointed out in order to use swarm in QIIME.

As Tomasz pointed out, use closed reference/open reference or denovo really depends on your question, what you want to look for and which type of data you're working with. The section 5.1.2 of the manuscript that Tomasz pointed out contains some directions on when to use one method or the other.

My recommendation, however, would be to start with closed-reference and check how many sequences you're loosing (in closed reference, the sequences that fail to hit the database are removed). If you're loosing too many sequences it will be worth to move to open-reference.

On the other hand, all this methods are based on OTUs, which has its own limitations. In general, we are trying to transition from OTUs to exact sequences, using denoising methods like DADA2 or deblur (which are currently not included in QIIME1.9.1, but will be in QIIME2).

Hope this helps!