PreprocessingMultiplexedIlluminaData

Introduction

This page will describe how to preprocess custom-barcode multiplexed Illumina data.

Prerequisites

You will need the following software in your executable path.

• TagDust - eliminate artifactual sequence from NGS data
• splitBC - a script to split barcoded reads. Found in ngopt repository under trunk/tools/splitbc
  If your reads are paired, you will also need the following program:
• repair - a program for pairing and shuffling Illumina reads. Found in ngopt repository under trunk/tools/pair_reads

You will need the following sequence files:

• adapter.fasta - Illumina adapter sequences.

• a tab-delimited file with your barcode sequences. This file should look something like this:

  BC1 ATGCTA
  BC2 TTGACT
  BC3 GTAGGT

• a single file containing all of your reads. They do not need to be shuffled if you have paired reads

Preprocess Multiplexed Data

Remove Artifactual DNA

Often times during library preparations, DNA fragments consisting only of adapter sequence are inserted and sequenced. To get rid of reads containing these sequences, we will use the program tagdust.

tagdust -o reads.clean.fastq -a reads.artifact.fastq adapters.fasta reads.fastq

where reads.fastq is our Illumina reads file. Reads without adapter sequence will be output to reads.clean.fastq

• TagDust has one key parameter, the false discovery rate (FDR). By default this is set to 0.01. This is the false discovery rate of classifying a read as coming from contaminant library sequences. (i.e. lower FDR, fewer discarded reads)

Split Barcoded Reads

To split our barcoded data set, we will use the splitBC from ngopt.

splitBC reads.clean.fastq --bcfile barcodes.txt --prefix ./ --suffix .fastq --bol

A file will be created for each line in barcodes.txt. For example, if we used the example barcode file from above, we would now have three files, BC1.fastq, BC2.fastq, and BC3.fastq

Re-pairing of Paired Reads

If you were processing paired-reads, your fastq files will likely contain broken pairings even if you shuffled them ahead of time. To fix this, you can use the repair program from ngopt.

repair -p ./ -s .fastq BC1 BC1.fastq

This will produce three files in the current working directory: BC1_p1.fastq, BC1_p2.fastq, BC1_up.fastq. The third file, BC1_up.fastq, will contain reads whose pair was discarded due to adapter contamination or barcode degeneracy.

If shuffled reads are desired, you can use the --shuf option for repair. BC1_shuf.fastq will be created instead of BC1_p1.fastq and BC1_p2.fastq

** To correct errors with Reptile, consider using the ReptileErrorCorrection tutorial **

** To assemble paired reads, consider using the FastQtoDraftAssembly tutorial. **

[Anonymous]

Originally posted by: Jacqueli...@gmail.com

Hello -- I want to use your repair program you have developed. However I am encountering a problem. I am not able to 'checkout' this program using the information stated on your source page. When I try to 'checkout' the program repair I receive this error message:

j?$ svn checkout http://ngopt.googlecode.com/svn/trunk/tools/pair_reads/repair.cpp svn: URL 'http://ngopt.googlecode.com/svn/trunk/tools/pair_reads/repair.cpp' refers to a file, not a directory

I would greatly appreciate it if you could assist me in downloading and installing this program. Also can you give me a more detailed description on how to use this program. Thank you for your time and assistance.