Improving assembly with PacBIo

[bioman1]

We have one paired-end (2x 101 bp)libray Whole gnome short sequence (WGS) sequenced through illumina hiseq 2000. We have assembled genome assembly with Abyss software. The assembly has lower N50 value around 1Kbp and does not improve in scaffolding because we have only one paired-end libarary. The estimated insert size is around 240 bp and coverage is arond 5x. Our sequenced plant is a polyploid, we lack data in getting good draft assembly and as we have budget constraint, we cannot go for additional illumina paired-end sequencing. But we some budget for 1x pacBio sequencing. My question does 1x pacBio sequencing helps in improving assembly atleast to get organelle genomes.

[gconcepcion]

We have one paired-end (2x 101 bp)libray Whole gnome short sequence (WGS) sequenced through illumina hiseq 2000. We have assembled genome assembly with Abyss software. The assembly has lower N50 value around 1Kbp and does not improve in scaffolding because we have only one paired-end libarary. The estimated insert size is around 240 bp and coverage is arond 5x. Our sequenced plant is a polyploid, we lack data in getting good draft assembly and as we have budget constraint, we cannot go for additional illumina paired-end sequencing. But we some budget for 1x pacBio sequencing. My question does 1x pacBio sequencing helps in improving assembly atleast to get organelle genomes.

With only 5x short read data, it is highly unlikely that 1x PacBio will improve any polyploid assembly, let alone a plant genome which is almost certainly highly repetitive.

If organellar genome is all you are interested in, you would be better off harvesting and concentrating the plastids using a cell sorter and subsequently sequencing the DNA extracted from the sorted organelles with PacBio

[rhall]

1x of PacBio will do absolutely nothing to improve the assembly, the minimum coverage would be ~10x for assembly improvement (gap filling / scaffolding). Even then, given that your illumina assembly is essentially not assembling (1kb N50 is significantly lower than a pacbio read length N50) it is highly unlikely that anything would be improved.
For a draft plant assembly there are no budget options, do you already have transcript data? it is generally much more cost effective for large genomes. Or as the previous poster points out, if you can reduce the problem (isolate organelles) and get higher coverage either for illumina or PacBio then an assembly will be much more successful.

[Chipper]

Well given the length of PacBio reads it would be hard not to improve the assembly even with 1x coverage since many of the small contigs will be attracted by single long reads. If your budget was calculated some time ago it might be that you can get substantially higher coverage given that it nov can give ~1 Gb per cell. Not saying that low coverage illumina + low coverage pacbio will give you a good assembly though.

[rhall]

The problem is that there is a low, but nonzero chance of a single read being chimeric (blunt end ligation of adapters can also result in blunt end ligation of fragments), therefore without multiple observations it is impossible to validate the long range information in a single read.

[bioman1]

Thank you for all replying. If I have 1x PacBio, does it help to map my illumina reads and I have some transcriptome dataset to this PacBio assembly and get organelle contigs (mitochondria and chloroplast). It will any form of information?