LoRDEC hybrid error corrected read usage
1
0
Entering edit mode
3.6 years ago
bio_d ▴ 20

Hi,

I am trying a donovo assembly of a reptilian genome (size comparable to humans) with ALLPATHS-LG. I have two illumina libraries paired-end and mate-pairs. In addition to it, I have a pacbio library.

I used LoRDEC to correct the errors in the pacbio data. For this I utilized the short reads from illumina (to get the deBruijn graph). I also carried out the trim-split step given in LoRDEC. My question is do I use the corrected pacbio reads (as is) or do I use the corrected-trimmed-split pacbio reads as long reads in ALLPATHS-LG deno assembly. pipeline

I am asking this because according to the LoRDEC manual "The output is the set of corrected reads also in FASTA format. In these corrected sequences: uppercase symbol denote correct nucleotides, while lowercase denote nucleotides left un-corrected."

Also, I plan to improve upon the correction process by using the corrected pacbio reads (either as corrected or as corrected-trim-split fasta files) as the input for the succeeding step of error correction with an increment in k-mer value and repeat the same. Could anyone tell me if the above steps are meaningful or if they are wrong, suggest an alternative iterated correction protocol.

Thanks

LoRDEC Pacbio Denovo ALLPATHS_LG • 1.9k views
ADD COMMENT
0
Entering edit mode

I have the same question as you. What did you end up doing? Trim, split, both or nothing?

ADD REPLY
0
Entering edit mode
3.6 years ago
Medhat 8.9k

The untrimmed un-split reads, contains uncorrected regions either because of lack of coverage or this regions of high errors and it could not be corrected. so using it could lead to miss-assemble.

regarding correction with different k-mer I think there is a suggested value by LoRDEC depends on the genome, as you mentioned it is a big genome, as I remember you should use 21.

Also I a recommend using HALC, based on this Efficiency of PacBio long read correction by 2nd generation Illumina sequencing

Regarding assembly: If you have high PacBio coverage (>20X) you can use canu for assembly without short reads.

also there is other ways to use PacBio

  1. filling gaps (case of low coverage)
  2. or hybrid assembly using tools like dbg2olc (Just an example, follow this link for more)

follow this post
C: Why we need 100X coverage to get a high-quality assembly?

ADD COMMENT
0
Entering edit mode

Thank you for the suggestions. The Pacbio coverage is approximately 6X, so I presume PBJelly might be a good option for me.

ADD REPLY
0
Entering edit mode

With this coverage it is a good option. Good luck

ADD REPLY
0
Entering edit mode
ADD REPLY

Login before adding your answer.

Traffic: 1762 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6