发信人: happyzhb (不再年青), 信区: Biology
标 题: Re: some silly questions about microarray
发信站: Unknown Space - 未名空间 (Thu Oct 28 09:44:05 2004) WWW-POST

There are primarily 2 types of microarray, comercial Affymetrix chip and
spotted array(cDNA/pcr), I guess u mean spotted array.
two major differences between the 2 types:
1. spotted material: affy use de novo synthesized oligo, 20mer or 25 mer,
called a probe, each gene is represented by 11? or 16 probes. the whole thing
called a probeset, so a probeset correspond to a gene. spotted array is
relative flexible, can print on the chip cDNA(comercially synthesized) or pcr
product, or even promoter sequence, whatever you want, for cDNA probes, length
is usually about 60~80mer.

in addition, correspond to each probe(perfect match(PM) to 25mer gene
sequence), there is 1 MM, meaning mismatch, intended to assess non-specific
hybridizations. Thus there are zillions of papers discuss how to use them.

2. affy chip: 1 chip 1 sample(1 hybridazation), spotted: 1 chip 2 sample(2
hybridizations with 2 dyes), so affy need to do inter-chip comparison and
normalization, spotted array need to ajust the influence of dye
incorporation(cy3 and cy5 incorportation efficiency in DNA are different)

So, now to ur question:
【 在 GreenPapaya (Turning Darkness into Light) 的大作中提到: 】
: 1. in each spot on the chip, how many sequences r fixed there? 1 or several?
Affy, no spot, but probe and probe set, 1 gene 11 or 16 probeset. Spotted
usually 1 gene 1 spot, but people who wants duplicates, will spot more spots
for 1 gene, depends on your preference, you control the robot, do whatever you
want.


: 2. how to degrade mRNA out after cDNA has been made?

Affy has its mannual, spotted array process vary amongst labs.

: 3. how can u ensure 1 chip includes all the possible gene sequences for the
: target cell genomes?

Affy tried its best, first affy human chip has 12,000 genes, basically
whatever annotated in NCBI is there, cannot do better.

spotted is customized, you want more, you spot more, if you only want genes
associated with one pathway, you can just spot 50-60 spots.

: 4. how can u detect if a same gene expresses more than once on the arrays?
how
: can u time it?

that's a sampling issue, you have to be able to time you sample lines.

: 5. there is a possibility that some cDNAs could not hybridize with the spot
: genen even they are matched, right? how can this be detected?

perfect match got to be hybridized. rarely happened, i guess, not sure about
this. don't need to detect them, the assumption is that global influence will
not change the whole picture.


Got go!

--
//昨天晚上我作了个梦，梦见我变成一只大蜜蜂
//嗡嗡嗡，嗡嗡嗡
//飞到西来飞到东

※ 来源:．Unknown Space - 未名空间 mitbbs.com．[FROM: 130.219.]