etheleon/pass repository overview

⁠PADI - Protein-guided Assembly and Diversity Indexing

⁠DOI:

⁠Table of Contents

• PADI - Protein-guided Assembly and Diversity Indexing⁠
  • Description⁠
    • Procedure⁠
      • Filters⁠
  • Usage⁠
    • Required Input⁠
  • Installation⁠
    • Docker Installation (Recommended)⁠
      • Example: SINGLE COPY GENES⁠
    • Installation from source⁠
      • 1. Dependencies/Pre-requisites⁠
      • 2. Install⁠
  • Description of pipeline⁠
  • Publication⁠
  • Future⁠

⁠Build

⁠Description

PADI short for Protein-guided Assembly and Diversity Indexing is the core module / tool in a pipeline for accessing diversity in complex microbial communities. It predicts number of genes within each KEGG (proteinaceous) gene family. PADI dynamically identifies regions of high diversity from contigs aligned to a guide consisting of reference protein sequences.

⁠Usage

maxDiversity    --megan </path/to/MEGAN> \
                --meganLicense <path/to/MEGAN5-academic-license.txt> \
                -f --outputDIR </path/to/output/dir> \
                --contigs ./example/data/contigs/ \
                --refseqKO ./example/refSeqProtDB/ \
                --threads 20

--contigs Folder containing binned contigs according to their KEGG family / orthology (NEWBLER 2.6 (20110517_1502))
--refseqKO Reference sequences grouped by their gene families 

⁠Procedure

1. BlastX of nucleotide contigs against reference prot sequences
2. Assign NTcontigs to best matched prot REFSEQ2.
3. MUSCLE generates MSA of reference prot sequences
4. Align NTcontigs to global REFSEQ MSA
5. Search for MAX diversity region

⁠Filters

• GAPS (ie. more gaps than are bases)

  • When the the gap ratio (NT:GAP) exceeds 1:10

• KOs where there are too few contigs are not considered

  • this or may not be IDEAL cause we’re looking for cases where there’s a large amt of RNA reads mapped to a low number o contigs

⁠Installation

⁠Docker Installation (Recommended)

Install⁠ Docker

⁠Example: SINGLE COPY GENES

cp SingleCopyGene SCG
mkdir SCG/out SCG/misc
#Place MEGAN5 license file inside misc
cp MEGAN5-academic-license.txt SCG/misc/

docker run --rm \
    -v `pwd`/SCG/data/konr:/data/refSeqProtDB \
    -v `pwd`/SCG/data/newbler:/data/contigs \
    -v `pwd`/SCG/out:/data/out \
    -v `pwd`/SCG/misc/MEGAN5-academic-license.txt:/data/misc/MEGAN5-academic-license.txt \
    etheleon/pass:0.1.2
    

⁠Installation from source

⁠1. Dependencies/Pre-requisites

• Tools

  • MEGAN5⁠
  • xvfb-run⁠
  • Muscle⁠. Use brew for installation.
  • Blast+

• Perl

  • v5.2X.X is required:
    • Use tokuhirom/plenv⁠ to install a local version of the latest perl (>=5.21.0)
    • Install CPANMINUS, using plenv install-cpanm

• R

  • ggplot2
  • dplyr
  • Biostrings (from Bioconductor)

⁠2. Install

cpanm https://github.com/etheleon/pAss.git

⁠Description of pipeline

The pAss pipeline requires one to provide contigs grouped by the their Ortholog groups. The scripts should be run in series from 00 to XX.

⁠Publication

In preparation

⁠Future

Include a sister software which uses protein HMMs on top of sequence similarity as a method to search for distantly related sequences.