Notable Tweets from Applied Bioinformatics & Public Health 2011

credits to http://pathogenomics.bham.ac.uk/blog/2011/06/all-the-tweets-from-abph11/

pathogenomenick: You will struggle to identify many species by 16S – particularly Streptococcus – even with full-length sequences #ABpH11

pathogenomenick: William Wade has a very nice database www.homd.org – 619 bact "species" represented, 66% of those cultured. 113 un-named. #ABPH11

pathogenomenick: Actinobacteria are underrepresented in 16S clone libraries: why – lysis? high GC? primers turned out to be the major problem #ABpH11

aunderwo: William Wade: Both bacterial culture and PCR amplification of 16S rRNA gene introduce their own biases when examining microbiomes #ABPH11

aunderwo: William Wade: By 16S rDNA sequencing found 50% of oral microbiome is unculturable #ABPH11

aunderwo: Whole transcriptome RNA sequencing will be possible with Ion Torrent 'soon' according to their rep #ABPH11

pathogenomenick: The Broad have done some early experiments with mate-pair sequencing on Ion Torrent using insert lengths of around 1.5kb #ABPH11
aunderwo: Mate paired libraries with 1.5kb inserts have been achieved with Ion Torrent PGM #ABPH11

aunderwo: At volume it is now possible to sequence a human genome for $4k using Illumina HiSeq #ABPH11
pathogenomenick: HiSeq 2000 – 8 human genomes per Tb or 8000 bacterial genomes per run. I know what I'd prefer! #ABPH11

avilella: Illumina HiSeq now: 600Gb per run. Latest R&D number: more than 1Tb per run #ABPH11

pathogenomenick: Super deep sequencing of KRAS allows detection of 1.1% variant frequency using MiSeq. This is going to take over cancer screening. #ABPH11

aunderwo: Possible to find SNPs involved in drug resistance in TB strains using MiSeq sequencing #ABPH11
jennifergardy: I'm switching my Christmas wish from a pony to an Illumina MiSeq. If they could throw in a pony w/ the machine, that would be great #ABPH11


jacarrico: Very nice examples of use for microbiology using miseq: TB, pseudomonas, ecoli sequencing #ABPH11
pathogenomenick: Presenting CF sputum metagenomics using HiSeq., PA LESB58 came out at 636x depth, plus phage & 7 other genomes 50-86% covered #ABPH11

pathogenomenick: After depleting human DNA, CF sputum DNA is >70% bacterial #ABPH11
jacarrico: Hiseq for metagenomics / Miseq to characterize individual isolates #ABPH11
fionabrinkman: MiSeq can seq P. aeruginosa LES genome accurately vs ref. which is good since large, high G+C. Using HiSeq for metagenomics though #ABPH11

avilella: The MiSeq pipeline will run the latest version of Velvet assembler as you can find at dzerbino's website. Nothing closed and canned. #ABPH11

fionabrinkman: @pathogenomenick yes! Next Star Trek movie must show shots of PacBio sequencing #ABPH11

pathogenomenick: OK, going to talk about Haiti cholera outbreak now. 12-fold genome coverage achieved in 90 minutes. Wow, those bugs are in log phase #ABPH11

aunderwo: FLX+ has a modal read length of 700bp – approaching read lengths of Sanger sequencing. Base accuracy 99.99%+ with 15x coverage #ABPH11
avilella: 454 FLX Plus modal 700bp, 85% above 500bp, total 700MB per run, 23 hours, accuracy a couple of 10^-5 extra pc points: 99,997% #ABPH11
jacarrico: 454 GS Flex + has 80% of reads greater than 500 bp and up to 1kbp. #ABPH11

lexnederbragt: RT @pathogenomenick: PacBio gives more even coverage of genome compared to Illumina – this is due to amplification bias. Models Poisson very well. #ABPH11

lexnederbragt: We too! MT @pathogenomenick: 454 8kb PE data can produce single scaffolds for S. pneumoniae, E. coli, (it's true, we've done it too) #ABPH11

aunderwo: Joo Andr Carrio: An ontology and REST API for microbial typing
Paper : http://bit.ly/jm7XDT Ontology: http://bit.ly/jzApXB #ABPH11 http://bit.ly/jm7XDThttp://bit.ly/jzApXB
pathogenomenick: RT @aunderwo: Joo Andr Carrio: An ontology and REST API for microbial typing
pathogenomenick: Has developed a RESTful MLST web interface. This is great. We just need it for next-gen now. #ABPH11

#ABPH11 http://rest.phyloviz.net/webui/

pathogenomenick: Developed data visualisation software called Phyloviz: http://bit.ly/isNJgj handles ST data, SNP data, looks pretty #ABPH11 http://bit.ly/isNJgj
pathogenomenick: .@jacarrico makes a compelling case for open data in molecular typing. What a shame it is not embraced by wider community #ABPH11
marina_manrique: Nick Loman @pathogenomenick starts the pipeline session. xBASE-NG A web interface for rapid analysis of bacterial genomes #ABPH11
aunderwo: Nick Loman: Web interface for WGS analysis http://ng.xbase.ac.uk/my/ #ABPH11 http://ng.xbase.ac.uk/my/

aunderwo: Nick Loman: Use Illumina sequence to correct homopolymeric tracts in 454 scaffolds #ABPH11
jacarrico: Nick Loman – illumina corrected 133 putative erros in 454 assembly #ABPH11
marina_manrique: Once more: the importance of hybrid assemblies (in this case #454 & #illumina) for correcting seq errors @pathogenomenick #ABPH11 #ngs

aunderwo: Marina Manrique: An annotation pipeline for NGS genome data http://www.era7bioinformatics.com/en/prokaryote_genome_annotation.html #ABPH11 http://www.era7bioinformatics.com/en/prokaryote_genome_annotation.html
jacarrico: www.ohnosequences.com – great name for a sequence assembler based on protein similarity #ABPH11

pathogenomenick: Mossong: Was alarmed when he got 1Gb files per each MRSA strain sequenced, compared with 7 bytes for MLST! #ABPH11
marina_manrique: Great! some info about the kind of technology used in Jel Mossong talk at #ABPH11 Illumina 2x100bp 80x for MSRA genomes
jacarrico: Interesting spa type vs SNP typing max parsimony tree comparison #ABPH11
aunderwo: Joel Mossong: using 85x coverage illumina data could extract MLST profiles from 36/40 strains. Puzzled about missing 4? #ABPH11
aunderwo: RT @pathogenomenick: Mossong: Was alarmed when he got 1Gb files per each MRSA strain sequenced, compared with 7 bytes for MLST! #ABPH11
marina_manrique: Another idea I've liked at Jel Mossong talk: WGS data should not be limited to SNP analysis, Mobile elements also play a role! #ABPH11

aunderwo: Marcus Claesson: comparing 454 and Illumina data for classifying bacteria using 16S. 454 outperforms Illumina #ABPH11
aunderwo: Marcus Claesson: Metagenomics – long reads of 454 give better classification , more data from Illumina => more OTUs detected #ABPH11
fionabrinkman: Claesson: 454 better vs Illumina for 16S seq (see http://goo.gl/u4CDD) but >60bp Illumina reads really helps & primer choice key #ABPH11 http://goo.gl/u4CDD)