From:
`To:
`Cc:
`Subject:
`Date:
`
`Richard Chen
`John West
`Scott Kirk; Christian Haudenschild; MarkPratt
`Re: Sprint 4 pre planning - JW questions & input
`Friday, August 10, 2012 9:14:52 AM
`
`let's talk sunday or monday john.
`
`the disk system is supposed to begin installation today I believe.
`
`On Fri, Aug 10, 2012 at 7:24 AM, John West <john.west@personalis.com> wrote:
`> The draft Sprint 4 list in Scott's email
`> below may include this (just in "short
`> hand") but it continues to be a top
`> accuracy priority to categorize the
`>>300k differences, when the same
`> genome is sequenced twice, by the
`> mechanisms which cause them, so we
`> can best prioritize potential solutions,
`> both bioinformatic and in the lab.
`
`>>
`
` Do we expect to have data back from
`> Illumina and Complete Genomics on
`> the CEPH1463 family and the Venter
`> genome ? If so, do we have a plan to
`> process and analyze them ?
`
`>>
`
` We should discuss lab priorities during
`> this sprint as well:
`
`>>
`
`CLIA development
`> Initial HiSeq runs (described in an email from Mark about 7/24)
`> Assay to resolve compressions etc using pulsed field electrophoresis
`> Development of Personalis-specific pull-out (carry over from Sprint 2)
`> Development of Personalis optimized methods for sequencing (in addition
`> to custom pullout)
`> Continued build-out of the lab facilities themselves
`
`>>
`
` As part of our exome effort, I wonder if
`> we should try processing an exome per
`> day in this sprint, again to get
`> experience and find what the normal
`> ranges are ? As I recall, there are a lot
`> of exomes in the SRA, some from
`> samples where we have already
`> processed the whole genome (good
`> to compare).
`
`>>
`
` Should we have an explicit goal to
`> get our new disk storage system
`> installed and working (storage
`> bottleneck removed) ?
`
`>>
`
` We can discuss more by email.
`> I should be back in the Bay area
`
`Personalis EX2077.1
`
`

`

`> late Saturday (now on a flight Cuzco
`> to Lima in Peru).
`
`>>
`
`>>
`
`>>
`
` Thanks,
`
` John West
`
`On Aug 9, 2012, at 7:01 PM, Scott Kirk <scott.kirk@personalis.com> wrote:
`
` Hi,
`
` Cell (408) 836-5586
`> jwest38261@aol.com
`
`>>>
`
`>>
`
`>>
`
` We're closing Sprint 3 tomorrow and will be planning Sprint 4 in the
`> afternoon. The areas we're targeting are in the list below. It consists of
`> the remaining work for prototype and will include some things carrying over
`> from Sprint 3.
`
`>>
`
` I had a chance to go through a first pass with Rich before he left and have
`> updated it somewhat to reflect where things are at currently.
`
`>>
`
` Irealize you're both out and we'll have to walk you through many of the
`> results when you get back next week. If you see this before then let me
`> know if you have any comments. It's going to be a stretch to finish what's
`> listed below in a sprint so we may need some help on priorities to stay
`> focused.
`
`>>
`
` Thanks,
`> Scott
`
`>>>>
`
`>>
`
`Some comments on Sprint 3:
`
`Here's what's on the list for Sprint 4 (in rough order of priority)
`
`>>>
`
`>>
`
` Exome support
`> Integration and testing work - we've incurred quite a bit of technical debt
`> and we should expect there to be some work in finishing integrating and
`> testing things so they function properly as a "pipeline".
`> Case/control dry run - Running a case/control “experiment” with a set of
`
` We're currently on day 12 of the 25 genomes in 25 days and have been working
`> through Java issues on the servers and what looks like some temperature
`> issues. We're at 9 genomes analyzed and have some fixes in place as of
`> today which could allow us to catch up a bit with more runs now in progress.
`> Annotation launcher has gone to 1.0 and now 1.1.
`> We have the variant statistics and at-a glance in place as well as a report
`> package design to demonstrate.
`> The tool updates are implemented and Hugo's working on testing them now, but
`> the server issues have been slowing that down.
`> The reports and madsen browning algorithm work remain on track
`> We're getting infrastructure in place for propagation of accuracy metrics
`> (likely won't finish by tomorrow though)
`
`Personalis EX2077.2
`
`

`

`> samples producing all outputs and packaging them up.
`
`>>
`
` significant blocks in getting data - waiting for dbgap approvals. If dbGap
`> comes through we'll need exome support for the data to use it. Short of that
`> we could make something up from the samples we have. This would allow us to
`> exercise all fo the functionality, but may not be the most scientific study.
`
` Completion of 25 genomes in 25 days effort
`
`>>
`
`>>
`
` started: 7/27, finish: 8/21
`> Calculate ranges
`
`>>
`
`>>
`
` tools testing
`> other sub-tasks from board
`
` Any items left from sprint 3:
`
` Personalis Report Package
`
`>>
`
`>>
`
` integrate the various pieces of the prototype design (with feedback)
`> Report explanations
`> automation
`> populate with ranges from 25 genomes work
`
`>>
`
` A/B comparison updates (indels)
`> Long indel detection (deletions) support
`> VNTR support
`> Addt’l ethnic specific reference support needs
`> BAMQC updates (if any)
`> These need to be done, but are probably lower priority
`
`>>
`
` CG post aligned data as input (if necessary, it may not be with use of the
`> CG converter)
`> Implement integration design for LIMS (we really don't have a system on line
`> yet to integrate with)
`
`Personalis EX2077.3
`
`