MakeMyTranscriptome (MMT): an automated pipeline for de novo transcriptomics

For a full guide, check out the Wiki

MMT is an attempt to automate a current best-practices transcriptome pipeline from start (read quality control) to finish (annotated transcriptome). If you provide metadata in a csv file, we will also do some basic pairwise differential expression with DESeq2. Since DE is not usually one-size fits-all, we recommend you look into these results and rerun as necessary.

MMT is composed of four principle tools:

• assembly - assemble input reads into a transcriptome
• quality - assess the quality of a transcriptome using common length-based and annotation-based metrics
• annotation - annotate a transcriptome using ORF-prediction and BLAST-style mappings
• expression - identify differentially expressed transcripts and genes

A fifth tool is currently being developed:

• compare - compare MMT assemblies constructed with different parameters

Each tool can be run individually, or all can be executed as a single virtual tool, full. All tools make use of two additional convenience modules, databases and tools which check for required databases and tools, respectively. If you're working on a Linux machine, MMT can download the tools for you; otherwise, it will assess the tools you have available and print recommended installation instructions.

One important note: if you already have a transcriptome, the full module can still be used to perform all downstream analyses. If you provide your transcriptome with the --assembly option, MMT will simply skip the assembly steps (currently this means skipping read QC as well, but this can be addressed if there's interest). Similarly, if you do not provide a csv file, MMT will skip the expression steps.

Quick Start Guide:

First, download MMT via git (below) or click the "Download ZIP" button.

git clone https://github.com/bluegenes/MakeMyTranscriptome.git

cd into the MakeMyTranscriptome directory.

cd MakeMyTranscriptome

MMT currently requires some python libraries that you should be able to install with anaconda or pip

• pandas
• Fadapa

To install the remaining tools and databases, run:

mmt setup --install --email youremail

If you just want to check the installed tools, omit the --install option. In release v0.1, databases will be downloaded in setup, but we'll update in a future release to allow you to specify a database path if you've previously installed these databases. Since downloading databases takes a while, we recommend using our --email option: MMT will email you with updates and when all databases have been installed.

We've included test data to check your installation:

mmt full -test --email youremail

A full test takes 20 minutes on 12 threads (13 mins of which is the assembly with Trinity).