Question

local fit and FourCseq

0

Entering edit mode

tlgolan • 0

@tlgolan-10804

Last seen 8.3 years ago

Hi,

Did anyone succeed to run the function getZscore while using local fit?

If so, please share.

local fit fourcseq • 1.3k views

ADD COMMENT • link updated 8.4 years ago by felix.klein ▴ 150 • written 8.4 years ago by tlgolan • 0

score 0 · Answer 1 · 2016-07-04

0

Entering edit mode

felix.klein ▴ 150

@felixklein-6900

Last seen 6.4 years ago

Germany

Hello, you need to use the parametric fit otherwise the getZScore function will not work because the estimated parameters are required. Best regards, Felix

ADD COMMENT • link 8.4 years ago felix.klein ▴ 150

0

Entering edit mode

Hi Felix,

Thanks for your quick answer.

I tried to use parametric fit, but a local fit was automatically substituted it.

How can I 'force' parametric fit to my data?

Thanks,

Tal

ADD REPLY • link 8.4 years ago tlgolan • 0

0

Entering edit mode

Hi Tal, this procedure is done automatically by DESeq2, so you cannot force the parametric fit. You can try to change the minCount = 40 threshold in the getZScores function so that the noisier part of the data is removed. Best regards, Felix

ADD REPLY • link 8.4 years ago felix.klein ▴ 150

0

Entering edit mode

Hi Felix,

When I chose minCount=372, I got a parametric fit. The problem is that this threshold left me with ~480 fragments only. I believe that this is a very low number of fragments. Is it enough to go on with the analysis?

Thanks again,

Tal

ADD REPLY • link 8.4 years ago tlgolan • 0

0

Entering edit mode

Hi Tal, this might indeed be to few fragments for the analysis. Did you have a look at the raw counts and how they are distributed? Also check the correlation plots with the different thresholds (after getZScores). How do they look? Best regards, Felix On 07/10/2016 03:03 PM, tlgolan [bioc] wrote: > Activity on a post you are following on support.bioconductor.org > <https: support.bioconductor.org=""> > > User tlgolan <https: support.bioconductor.org="" u="" 10804=""/> wrote > Comment: local fit and FourCseq > <https: support.bioconductor.org="" p="" 84617="" #84878="">: > > Hi Felix, > > When I chose minCount=372, I got a parametric fit. The problem is that > this threshold left me with ~480 fragments only. I believe that this > is a very low number of fragments. Is it enough to go on with the > analysis? > > Thanks again, > > Tal > > ------------------------------------------------------------------------ > > Post tags: local fit, fourcseq > > You may reply via email or visit > C: local fit and FourCseq >

ADD REPLY • link 8.4 years ago felix.klein ▴ 150