## User: Moshe Olshansky

Moshe Olshansky •

**250**- Reputation:
**250**- Status:
- Trusted
- Location:
- Last seen:
- 7 years, 1 month ago
- Joined:
- 8 years, 6 months ago
- Email:
- o********@wehi.edu.au

#### Posts by Moshe Olshansky

<prev
• 25 results •
page 1 of 3 •
next >

0

votes

1

answer

757

views

1

answers

... Hi Paolo,
Your command:
colnames(design) <- c('CME','ES','CMA','CMN')
is wrong (i.e. this is not what model.matrix produces).
Do the two previous commands, i.e.
sampleType <-
c('CME','ES','CMA','CMN','CME','ES','CMA','CMN','CME','ES','CMA','CMN'
)
design <- model.matrix(~ factor(sample ...

written 7.1 years ago by
Moshe Olshansky •

**250**0

votes

1

answer

2.0k

views

1

answers

... Hi Ingrid,
If I understand correctly, you would like to find genes which are
differentially expressed (DE) between Treatment and Control at 4 hours
and
compare them with those which are DE at 18 hours.
One way to do it is to split your data into two separate sets ( 4
hours
and 18 hours) and find DE ...

written 7.2 years ago by
Moshe Olshansky •

**250**0

votes

1

answer

2.0k

views

1

answers

... Hi Ingrid,
With your design your "base" level is patient 4, Control, 4 hours
(let's
call it B).
The mean for, say, patient 6, Treatment, 18 hours is:
B + Donor6 + TreatT + Time18
where Donor6 is the difference between Donor4 and Donor6 (same for any
treatment and time), TreatT is the difference bet ...

written 7.2 years ago by
Moshe Olshansky •

**250**0

votes

1

answer

687

views

1

answers

... Hi Rich,
You have already got the answer from Kasper. This is exactly what I am
suggesting.
The idea is that after log transformation the variances of the genes
follow some distribution. So the more genes you are using the better
you
can estimate that distribution. This is just a model, nobody is
c ...

written 7.3 years ago by
Moshe Olshansky •

**250**0

votes

1

answer

687

views

1

answers

... Hi Rich,
I think that Gordon Smyth (the author of limma) has explained at this
list
what moderated t-statistic is.
The brief explanation is that when there are few samples the estimate
of
the variance which is used in a standard t-test is quite noisy and
because
one must account for this noise the ...

written 7.3 years ago by
Moshe Olshansky •

**250**0

votes

1

answer

687

views

1

answers

... Hi Rich,
Whether to use the moderated t-statistic or not does not depend on
whether
you are interested in the 10 particular genes or in all differentially
expressed ones. This will affect your multiple testing adjustment.
The simplest way for you to proceed is to use limma as usual, get the
topTabl ...

written 7.3 years ago by
Moshe Olshansky •

**250**0

votes

1

answer

456

views

1

answers

Answer:
A: Random Forest Code

... Here is the link to the original Fortran 77 code:
http://www.stat.berkeley.edu/~breiman/RandomForests/cc_examples/prog.f
See also
http://www.stat.berkeley.edu/~breiman/RandomForests/
The R package should probably contain the wrapper code (but ask the
package creator).
Best regards,
Moshe.
> D ...

written 7.3 years ago by
Moshe Olshansky •

**250**0

votes

2

answers

1.1k

views

2

answers

... You are right. My mistake. Sorry...
> Hi, Moshe,
>
> I still did not get it. How to use your method?
>
> lmFit<-(x,design) which x allow missing values, but design didn't.
>
> usually, I want to adjust batch effects, I will include it in design
> matrix, i.e.
> ...

written 7.4 years ago by
Moshe Olshansky •

**250**0

votes

1

answer

476

views

1

answers

... You can definitely use cyclic loess normalization for one channel
arrays
unless your experiments are very different, in which case any
normalization can be problematic.
Moshe.
> Dear all,
>
> Is it plausible to do cyclic loess normalization on single channel
> arrays coming from differ ...

written 7.4 years ago by
Moshe Olshansky •

**250**0

votes

2

answers

1.1k

views

2

answers

... Why do you want to include weight in the design matrix?
It may be more reasonable to include weight in the linear model, i.e.
expression ~ condition + weight and then your design matrix will have
two
columns (if there are two conditions): the first column will contain
0's
and 1's depending on the co ...

written 7.4 years ago by
Moshe Olshansky •

**250**#### Latest awards to Moshe Olshansky

No awards yet. Soon to come :-)

Use of this site constitutes acceptance of our User
Agreement
and Privacy
Policy.

Powered by Biostar
version 16.09

Traffic: 125 users visited in the last hour