| layout | root | title | date |
|---|---|---|---|
lesson |
. |
mothur tutorial |
2017-08-04 |
Instructor: Tracy Teal [email protected] @tracykteal
What you can use it for and what is the workflow in slides:
http://tracykteal.github.io/mothur-tutorial/mothur_presentation_2017.pdf
You can run mothur on your own computer or an a remote computer. For this exercise we'll be running it on the MBL servers.
Type
mothur
Now you're in mothur, and we can start!
You should get the command prompt
mothur >
Green sticky note up if you have the mothur command prompt
Once you have it up the first thing we'll do is quit mothur, so type
quit()
If you're in mothur, or really any command line environment, you can type Ctrl^C to quit that command
Getting help
mothur manual - http://wiki.mothur.org/wiki/Mothur_manual
mothur user forum - http://www.mothur.org/forum/
How are you going to keep track of all the commands you're running.
Be nice to the you of 6 months or even a week from now!
Computational processes need lab notebooks just like lab work.
- Evernote
- OneNote
- Word
- a text editor
- IPython notebook
- photographic memory
In mothur if you run an analysis with different parameters in the same directory, it will write over your old analysis.
mothur does always write out logfiles, so you can keep track with those some. We'll look at those after we've run some commands.
Let's take a look at the data we have and how it's organized.
cd mothhur_data
ls
We have all our files in 'MiSeq_SOP'. There are the FASTQ files from the experiment, as well as some reference files we need for the analysis. We will end up with a lot of files in this one directory. We'll organize things a bit at the end.
The file that you need to create before any analysis, is a list of the samples and the sequences associated with those samples. Here it's called stability.files but you can call it anything. It looks like this.
You can see that the format is a tab delimited file with the sample name in the first column, the forward read in the second column and the paired end read in the third column.
(Optional: look at the stability.files file with 'less')
Working with mothur, we're going to follow the mothur MiSeq SOP.
The command we'll use on the MBL server are here.
Follow along on this document: MBL mothur MiSeq SOP
Once we create the shared file, there are several things we can do. First we can download this file to our local computers.
Now let's take a look at the shared file. Let's open it in Excel. stability.trim.contigs.good.unique.good.filter.unique.precluster.pick.pick.pick.an.unique_list.shared
Let's also look at the list file
stability.trim.contigs.good.unique.good.filter.unique.precluster.pick.pick.pick.an.unique_list.list
There's been a lot of discussion of what to do with the raw counts:
- subsample
- normalize by sample size
- mixed model
- something else
For this lesson we have subsampled. When we subsample, things will look different each time. Compare your 'shared' file with that of your neighbors. Are they the same?
Let's say that instead of using the 0.03 cutoff when you make your shared file, you want to use 0.05?
By default if you run the same command with a different parameter it just overwrites the previous files. Try running
make.shared(list=stability.trim.contigs.good.unique.good.filter.unique.precluster.pick.pick.pick.an.unique_list.list, count=stability.trim.contigs.good.unique.good.filter.unique.precluster.uchime.pick.pick.pick.count_table, label=0.05)
and see.
Rerun it again with 0.03 to get it back to where it was.
Instead you can redirect the output to a new folder with set.dir. Let's try it. We'll tell mothur to use the input directory 'analyses/first' where we have all the files that have been generated already, then we'll set a new output directory.
system(mkdir ~/data/second)
set.dir(input=~/data/process, output=../data/second)
Then run
make.shared(list=stability.trim.contigs.good.unique.good.filter.unique.precluster.pick.pick.pick.an.unique_list.list, count=stability.trim.contigs.good.unique.good.filter.unique.precluster.uchime.pick.pick.pick.count_table, label=0.05)
Ta da, the 0.05 file is now in 'second'. You can do this sort of thing anywhere along the path. Instead of explicitly setting the output and input, you can also point to an input file directly.
make.shared(list=~/data/process/stability.trim.contigs.good.unique.good.filter.unique.precluster.pick.pick.pick.an.unique_list.list, count=~/data/process/stability.trim.contigs.good.unique.good.filter.unique.precluster.uchime.pick.pick.pick.count_table, label=0.05)
Once you've run through mothur with your shiny new data, you're get to a point where you're happy with a certain set of parameters. Now you might want to set mothur up in batch mode. That's where you make a file with a bunch of commands and then just run that file.
We have one of those files already in our data directory: stability.batch
When you're doing this the fact that mothur remembers what files you used last is great, because you don't have to type it all out.
Let's try making our own. We'll go back on the server and use nano.
Put the first couple of commands we run in to a file, along with comments about what we are doing.
#set input and output directory
set.dir(input=~/Desktop/mothur-project/data/MiSeq_SOP, output=~/Desktop/mothur-project/analyses/first)
#make contigs
make.contigs(file=stability.files, processors=8)
We're using the 'full path' of the file names here so there's no confusion.
You have been working in R already, so there's a lot of ways you can work with the data there.
There is also PAST - a GUI statistical package developed by Oyvind Hammer.
http://folk.uio.no/ohammer/past/