RobertsLab / resources

https://robertslab.github.io/resources/
19 stars 11 forks source link

estimate cost for BS sequencing of Salmon Samples #678

Closed shellywanamaker closed 5 years ago

shellywanamaker commented 5 years ago

Complete cost analysis for different methods (RRBS, WGBS, and Amplicon-based BS-seq) and include it here:

https://docs.google.com/presentation/d/1mD6ivb3xcfhvu_fF7dhYTLWytzXeYp8rheEvbTJKWnc/edit#slide=id.p

kubu4 commented 5 years ago

@yaaminiv @sr320 I think this should be merged with this issue:

https://github.com/RobertsLab/resources/issues/543

as costs will be/are part of the decision tree.

sr320 commented 5 years ago

Seems like there could be a universal google sheet where you input species info and everything calculates. Maybe a Shiny app one day, but lets focus on getting some numbers first!

thanks Steven On Apr 11, 2019, 2:59 PM -0700, kubu4 notifications@github.com, wrote:

@yaaminiv @sr320 I think this should be merged with this issue:

543

as costs will be/are part of the decision tree. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

kubu4 commented 5 years ago

Seems like there could be a universal google sheet where you input species info and everything calculates. Maybe a Shiny app one day,

Whoa! Killer idea! Hackweek 2019?

shellywanamaker commented 5 years ago

Here's an update on where we're at with pricing: Sea lice: using swift (accel) library prep,an estimated 240M reads would get >90% coverage of all mCpGs at 10X

Quotes Sam got: Zymo: $2300/sample (i.e. $4600 for the project). Genewiz: $2320 ($880/sample + $1440 for 1 lane of sequencing; Hollie thinks they use the Swift Accel kit)

Salmon:
The 2017 Scientific Reports paper on temp stress methylome and transcriptome response in Atlantic Salmon used the Zymo Methyl-miniseq protocol. Zymo quote for the following services (sample QC and standardization, library construction, Next-Generation Sequencing on Illumina Hiseq, and raw data alignment; prices include a 10% discount for educational institution and 5% discount for bulk order- 12 or more samples): Methyl-miniseq: $1105/sample Classic RRBS: $551/sample more info about services

To do library prep in house: ~$750 not including adapter or sequencing cost (for all samples) Taq-alpha I = $250 MspI = $100 Zymo DNA concentrator columns (150) $150 Illumina bisulfite seq adapters Agarose gel and gel clean up kit (qiagen QG buffer would do, $50; could alternatively do bead cleanup) Bisulfite treatment kit (EZ DNA Methylation-Lightning Kit) $200 PCR reagents

*could substitute these steps with the Swift Accel NGS methyl-seq kit ($1750 for 24rxn kit and $120 for indexing kit; prices may be lower if we ask for an edu. institution quote); this kit really only outperforms traditional methods when DNA input is low)

*could also substitute these steps with NEBnext Enzymatic Methyl-seq kit. List price: $895 for 24 rxns.

shellywanamaker commented 5 years ago

1) For all 20 salmon samlpes: low end is ~$6,000: RRBS MSPI/Taq-alphaI digest and clean up ($150) + NEB EM-seq kit ($895) + 1600M reads on UW's Novaseq ($4714).

2) To do the same thing with an in-house kit is ~$6,300: RRBS MSPI/Taq-alphaI digest and clean up ($150) + home brew kit ($1460) + 1600M reads on UW's Novaseq ($4,714).

3) Prep with Swift Accel NGS methylseq Kit ~$6,700: RBS MSPI/Taq-alphaI digest and clean up ($150) + Swift Accel NGS methylseq Kit ($1,870) + 600M reads on UW's Novaseq ($4,714).

4) Zymo methylmini-seq service $22,100 .

This is based on this info: https://docs.google.com/spreadsheets/d/11OWceVZ9pehSSwnR_WjBQncOkRHxNLyZAR2BRNu5aVw/edit?usp=sharing , this github issue, this github issue, and https://docs.google.com/presentation/d/1mD6ivb3xcfhvu_fF7dhYTLWytzXeYp8rheEvbTJKWnc/edit#slide=id.g50f7fba360_0_22.

If we want to play it safe, probably do the MSPI/Taq-alphaI digest (RRBS MethylMini-seq protocol) done in the The 2017 Scientific Reports paper on temp stress methylome and transcriptome response in Atlantic Salmon followed by our in-house library prep kit and sequencing at UW on the Novaseq.

@sr320 what do you think?

sr320 commented 5 years ago

@shellytrigg - so you recommend option 2?

shellywanamaker commented 5 years ago

@sr320 yes, if we want to follow what the Sci Reports paper did. If the gDNA yields end up being low, we can re-evaluate

shellywanamaker commented 5 years ago

A little change up, the plan is to:

  1. Isolate DNA using the EZna extraction kit (already have in lab)

  2. RRBS MSPI/Taq-alphaI digest and clean up

  3. Library prep with Zymo Pico Methyl-seq kit (because this is the kit we are using for the sea lice samples, it will be easier to prepare everything the same way). I confirmed with Zymo that the kit should work fine with RRBS digested samples because it uses their gentlest bisulfite conversion (EZ lightening) ***as long as the UW sequencing core uses ONLY Illumina products for sequencing (no homemade reagents). I also confirmed the kit is compatible with NovaSeq flowcell technology and it's possible to multiplex all our samples (see email about PicoMethylSeqKit_IndexAlternatives), we just need to order ~17 oligos (1 universal + 16 with unique indexes; ~10$/60bp oligo so ~$170 total)

  4. Sequence on NovaSeq 6000 with SP flowcell for 1600M reads (240M for sea lice 1, 240M for sea lice 2, and 56M/salmon sample).