Prost!

Prost! (PRocessing Of Short Transcripts) can analyze smallRNA sequencing data generated on any sequencing platform. Prost! does not rely on existing annotation to filter sequencing reads but instead starts by aligning all the reads on a user-provided genomic reference, allowing the study of miRNAs in any species. Additionally, any number of samples can be studied together in a single Prost! run, allowing an accurate analysis of an entire dataset. After grouping the processed reads by genomic location, Prost! then annotates them using a user-defined annotation database (public or personal annotation database). Genomic alignment, grouping, and then annotation enable the study of potentially novel miRNAs, as well as permitting the retention of all the isomiRs that a miRNA may display. Finally Prost! contains additional features such as grouping by seed sequence for a more functional approach of the dataset, provides automatic discovery of potential mirror-miRNAs, and analyzes the frequency of various types of post-transcriptional modifications at each genomic location. Each step of the Prost! analysis is provided in an Excel output file so that the user have access to all information for deeper analysis of specific cases.

Authors

  • Peter Batzel - Conceptual Design, Software Engineering, and Algorithm Design
  • Thomas Desvignes - Conceptual Design Leader
  • Jason Sydes - Conceptual Design, Software Engineering, and Algorithm Design
  • Brian F. Eames - Prototype Design
  • John H. Postlethwait - Project Advisor

News

March 8th, 2019: Prost! was published in Scientific Reports.

Site Contents

In addition to this home page, there are several other pages on this site as well. Click on the Site link in the header to access those pages. Here is an overview of the content of the other pages.

Installation of Prost!

Installation via pip is probably the easiest way to get Prost! installed. To install to your home directory:

pip install prost --user

Or you can install systemwide (if you have sudo privileges):

sudo pip install prost

Alternatively, you can download a release and install directly with setup.py:

# Point your browser at
https://github.com/uoregon-postlethwait/prost/releases/latest

# And use your browser to download the latest release.  For example:
https://github.com/uoregon-postlethwait/prost/archive/v0.7.3.tar.gz

# Extract
tar xzf prost-0.7.3.tar.gz
cd prost-0.7.3

# Install (systemwide)
python setup.py install

# Install (in your home directory)
python setup.py install --user

Finally, you can clone the git repository and install directly with setup.py:

# Clone repository
git clone https://github.com/uoregon-postlethwait/prost
cd prost

# Install (systemwide)
python setup.py install

# Install (in your home directory)
python setup.py install --user

Quickstart

To help you get Prost! up and running quickly, we’ve provided a demo in the style of a tutorial. The demo includes all the sample data you’ll need (except for the reference genome due to size constraints, which you’ll have to download separately).

To get started, see the Quick Start Demo page.

Requirements and Dependencies

Prost! is designed to be easily run from the command line. A basic familiarity of the command line and optionally the ability to edit a configuration file are all that is required to Prost!.

Listed below are the various requirements that need to be fulfilled in order to run Prost!.

System and software dependencies

  • A Linux or Mac OS X environment.
  • Python version 2.7.x (tested with Python 2.7.9, 2.7.5, and 2.7.2).
  • BBMap short read aligner, available for download at SourceForge. See Installing BBMap.

Configuration files

  • The Prost! configuration file (default prost.config). We recommend you download our example Prost! configuration file (prost.config.example) and modify it to suit your experiments. The example configuration file is well documented and should be easy to quickly comprehend and edit.
  • A list of high throughput sequencing FASTA format files of samples (default samples_filelist). The format is fileName descriptiveName, with one samples file per line.

Samples FASTA files

  • Preprocessed input fasta files containing short sequence reads from your samples. Quality filtered and barcodes or adapters removed so that the only sequences remaining in the files are full, high quality, short sequences.

Reference genome BBMap database

Prost! requires a BBMap database of a reference genome of your species of interest (or a closely related species). The reference genome does not need to be annotated or assembled to high N50s. We have tested Prost! on PE250 Illumina sequencing reads on a non-model organism with good results. See Building BBMap Databases for instructions on building a BBMap database.

Annotation FASTA files

Three annotation FASTA files are required by Prost!. (NOTE: At this time, nucleotides uracil and thymine must be coded as T’s (and not as U’s) in the annotation FASTA files.)

  • A FASTA file of annotated mature microRNAs. This is usually built directly from miRBase’s “mature.fa” (or from an augmented version of that FASTA file), and should contain mature miRNAs from several species (including the species you are studying). The FASTA header file must follow miRBase’s convention of prefixing each miRNA with a three letter species abbreviation and a dash. For example:

    >dre-miR-451 MIMAT0001634 Danio rerio miR-451
AAACCGTTACCATTACTGAGTT

  • A FASTA file of annotated microRNA hairpins. This is usually built directly from miRBase’s “hairpin.fa” (or from an augmented version of that FASTA file), following the same conventions for the mature miRNAs described above.

  • A FASTA file of annotated “other RNAs” (e.g snoRNAs and lincRNAs) for the species under study, or a related species. See Generating a BioMart Other RNA annotations BBMap database. Following those instructions will produce correctly formatted FASTA files (the FASTA header format is: >geneName|biotype|geneID). Prost! requires a BioMart file because it usually provides useful annotations. If you do not wish to supply a real BioMart file, simply download and point Prost! at the following fake BioMart file: fake_biomart_file.fa

Prost! Output

See Prost! Output for a description of the output of Prost!.

De novo miRNA annotation

See De novo miRNA annotation to read about a method of using Prost! output to perform de novo miRNA annotation.

Command Line Help

Prost! provides configuration options provided through the command line. To see a list of available options with contextual help, as well as each options defaults, type the following in a terminal window:

prost -h

# or
prost --help

Configuration

Prost! follows the usual convention of allowing configuration via defaults, a configuration file, and command line flags. Sensible defaults are provided, so in general very few configuration options will be needed by the user.

A minimal configuration file (by default named prost.config) is required by Prost! (you can download and modify our example configuration file here: prost.config.example). At minumum, the configuration file needs to specify the species under study (e.g. ‘mmu’ for mouse) in the General section. In addition, the GenomeAlignment section needs to specify the following:

  • name: A user-defined name for the alignment (default: genome).
  • tool: The alignment tool being used (default: bbmap; currently only BBMap is supported).
  • db: The BBMap database of the reference genome (e.g. a BBMap database) to be used for the Genome Alignment. (Note that for the AnnotationAlignment sections below, this db field must point to a FASTA file instead of a BBMap database!)
  • max_3p_mismatches: The maximum number of mismatches allowed for each sequence alignment hit on the 3´ end of the hit (default: 3)
  • max_non_3p_mismatches: The maximum number of mismatches allowed for each sequence alignment hit anywhere but the 3´ end of the hit (default: 2) (i.e. either on the 5´ end or in the middle of the sequence).
  • allow_indels: Whether or not to allow alignments which have insertions or deletions.
  • indelnt_penalty_multiplier: Multiply the number of inserted or deleted nucleotides in the alignment by this penalty, and add that to the number of 3´ mismatches when determining whether to keep or reject that alignment. Basically, this can nearly eliminate indels from your dataset if that is your desire.

Below is an example of a minimal configuation file:

[General]
species: mmu
samples_filelist: data/samples_filelist

[GenomeAlignment]
name: genome
tool: bbmap
db: /path/to/databases/Danio_rerio.GRCz10.dna.toplevel
max_3p_mismatches: 3
max_non_3p_mismatches: 2
allow_indels: yes

The configuration file as described above will not perform any annotations. See the file prost.config.example for a working example of annotation alignment sections. The AnnotationAlignment sections follow the same structure as the GenomeAlignment section; the only difference being that the db field must point to a FASTA file instead.

Currently, (nearly) every option in the General section of the configuration file can also be controlled via command line flags. See Command Line Help.

Running Prost! On Your Own Dataset

To run Prost! on your own dataset, you’ll need to edit the configuration file. In particular, you’ll need to edit the fields ‘species’ and ‘samples_filelist’ (in the General section) and the ‘db’ fields (in the Alignment sections). Below is a snippet of the configuration file that shows you roughly what will need to be edited (some additional fields are shown below for context, but do not need to be edited):

[General]
species: dre
samples_filelist: samples_filelist
mature_mir_annotation_fasta: BBMap/mature_miRBase21.fa

[GenomeAlignment]
db: BBMap/Danio_rerio.GRCz10.dna.toplevel

[AnnotationAlignment1]
type: MirbaseMirAnnotation
db: BBMap/mature_miRBase21.fa

[AnnotationAlignment2]
type: MirbaseHairpinAnnotation
db: BBMap/hairpin_miRBase21.fa

[AnnotationAlignment3]
type: BiomartOtherRNAAnnotation
db: BBMap/BioMart_Dre89_otherRNA.fa

After you have made those changes, simply run Prost! again:

prost

Funding

Prost! has been funded by the following grants:

  • Identification of MiRNAs Involving Midfacial Development and Clefting; NIH - National Institute of Dental and Craniofacial Research (U01 DE020076)
  • Advancing the Scientific Potential of Transcriptomics in Aquatic Models; NIH - Office of the Director (R24 OD011199)
  • Resources for Teleost Gene Duplicates and Human Disease; NIH - Office of the Director (R01 OD011116)
  • Mechanisms of Sex Determination in Zebrafish; NIH - National Institute of General Medical Sciences (R01 GM085318)
  • Developmental Mechanisms for the Evolution of Bone Loss; NIH - National Institute on Aging (R01 AG031922)
  • Signaling Hierarchies in Vertebrate Development: CP1: A zebrafish model of phenotypic variation associated with Fraser syndrome; NIH - Eunice Kennedy Shriver National Institute of Child Health and Human Development (P01 HD22486)
  • Antarctic Fish and MicroRNA Control of Development and Physiology; NSF - Office of Polar Program (OPP #154338)

Citing Prost!

If you use Prost!, please cite it like so:

Thomas Desvignes, Peter Batzel, Jason Sydes, B. Frank Eames, and John H. Postlethwait. 2019. “MiRNA Analysis with Prost! Reveals Evolutionary Conservation of Organ-Enriched Expression and Post-Transcriptional Modifications in Three-Spined Stickleback and Zebrafish.” Scientific Reports 9 (1): 3913. https://doi.org/10.1038/s41598-019-40361-8.

Prost! has been previously cited like so:

Thomas Desvignes, Peter Batzel, Jason Sydes, B. Frank Eames, & John H. Postlethwait. (2019, March 9). Prost!, a tool for miRNA annotation and next generation smallRNA sequencing experiment analysis (Version v0.7.53). Zenodo. http://doi.org/10.5281/zenodo.594353.