Best parameter for bacteria de novo assembly - HGAP

[heyyyjude]

Hello,

I would like to know what is best for bacteria genome assembly in general. I work on 4M~6M genomes in general and they have 1 or 2 chromosomes with some plasmids. But I have more contigs than I expected them to be, so I would like to get reduced contigs.

I leave all parameters by default except genome size and I use HGAP3. So I am wondering if there are more options get reduced contigs, apart from genome size and using HGAP2. raising Target coverage or any other parameters help to get reduced contigs?

I work with PacBio20K and after running SMRT analysis (HGAP3) I usually got 80x for a chromosome but sometimes I got more than 100x for a chromosome.

Sorry about my English if it bothers you.
any advice would be appreciated.
Thanks.

[rhall]

HGAP.3 is very robust, trying to optimize parameters for a bacterial assembly is not going to have much of an effect. What is the subread N50 for your data, and the seed length used in the preassembly? I would suggest trying to understand why the assembly is not resulting in complete chromosomes, try calculating dot plots between the contigs to see if the reason for the breaks is repetitive sequence. Or try running the Bridgemapper protocol on the HGAP output to see if you do have evidence for joining any of the contigs.

By far the best way to improve a HGAP assembly is to add more longer reads, either by increasing sequencing depth, or generating an improved library.

[heyyyjude]

Thank you for you advice. I will try to understand why it's not completed as you said.