PacBio performance and efficiency in the field

[Trev]

I have been searching for some down to earth information regarding how the PacBio RS is actually performing that doesn't come from PacBio marketing in terms of both the wet bench side as well as the data quality coming off the system. Is there a honest review of the system out there yet? The system has been out there for awhile now.

For instance, how is the system performing? I was at a seminar where a sales rep publicly indicated they were having variability problems.

How many SMRT cells are actually required to not only produce whole-genome sequence data for say an E. coli BUT to produce enough and good enough sequence that can be assembled? The reason I ask is that it looks like from the PacBio publication on the O104 outbreak that it took >50 SMRT cells (cells for both closed circular and long subreads) to assemble the single outbreak isolate. Does it really take this many to get an assembly with PacBio data? Why collect so many if it doesn't?

What is the final cost of getting assembly quality whole-genome sequencing data off of the PacBio given the number of SMRT cells that need to be run?

Is the best use of the PacBio RS for whole-genome sequencing in combination with other sequencing data of higher quality and throughput like an Illumina system?

Is anyone using the PacBio RS for whole-genome sequencing of genomes larger than microbes?

Is there any public review on the web that lists this information that anyone can link to or comment on? It seems like it is slim pickings out there for actual information.

[wweise]

PacBio information from customers

Full disclosure, I am from the marketing department at Pacific Biosciences but didn't want your question to go unanswered. We have just launched several application-focused pages on our website with customer results, including: webinars, technical posters, peer-reviewed publications, etc. Please visit: www.pacb.com/apps. I hope this helps answer some of your questions....

[petang]

I am using PB to do de novo genome sequencing of an ~170MB genome.
The followings are the results form 4 SMRT cells:
% Adapter Dimer (0-10bp) 3.96
% Short Insert (11-100bp) 0.51
Pre-Filter # of Bases 985171659
Post-Filter # of Bases 537023600
Pre-Filter # of Reads 601168
Post-Filter # of Reads 132250
Pre-Filter Mean Readlength 819
Post-Filter Mean Readlength 3393
Pre-Filter Mean Read Quality 0.191
Post-Filter Mean Read Quality 0.842

Therefore, around 22% of the reads are QUALITY reads.
The mean read length of 3Kb is good enough though.

I will to assemble the datasets and update the result later.