l****u 发帖数: 46 | 1 http://www.ncbi.nlm.nih.gov/sra
Sequence Read Archive (SRA) and Trace Archive repositories have been
discontinued
Due to budget constraints, NCBI will be discontinuing its Sequence Read
Archive (SRA) and Trace Archive repositories for high-throughput sequence
data. Closure of the databases will occur in phases. SRA and Trace will stop
accepting some types of submissions in the coming weeks, and all
submissions within the next 12 months. Over the next several months, NCBI
will be working with sta... 阅读全帖 |
|
j*****d 发帖数: 787 | 2 genetics的人看问题很难到那个深度,他们只会去decode A T C G
sunny的文章读起来有点像我早年读那个违反中心法则的nature文章那样
读了让人身上起冷汗(没有任何恶意,只是觉得太复杂)
——有些genetic locus的向下传递 复杂程度,可能的确超乎我们的想象。 |
|
j*****d 发帖数: 787 | 3 恩 我孤陋寡闻
大牛贴些连接给我瞧瞧
我最近碰到一个genetic locus很复杂,promoter 怪异,transcript多,翻译后修饰多
蛋白, mRNA表达也与功能 机密调节 |
|
l**********1 发帖数: 5204 | 4 Both are new 2012 Nature papers:
//www.nature.com/nature/journal/vaop/ncurrent/full/nature10945.html
尤其这篇
//www.nature.com/nature/journal/vaop/ncurrent/full/nature11011.html
its full text noted SHANK3:
//www.hudsonalpha.org/sites/default/files/nature11011.pdf
NB: 不温故SHANK3哪来知新 SNP
还有要是不同组出了 自闭症 自相矛盾的gene locus SNP 结果一次的话 还好说 注意这里指基因名称
反复出现的话
Nature Senior Editor 2011 还混2013 senior editor 那 啊!?
|
|
n******7 发帖数: 12463 | 5 学艺不精,一直没搞清楚。如果一个locus上的两个allele,分别是显性的A和隐性的a
,他们在分子水平的对应物是什么?
DNA水平的话,现在用的reference genome sequence是单倍体的(除了chr6等一些区域
),也就是说参考的DNA seq都是一样的。如果两个等位基因DNA sequence不一样,
reference sequence用哪个?
RNA水平的话,那就是isoform了?但是DNA sequence都一样的话,两个allele被调控起
来也是一样的,怎么产生不同isoform来对应不同的表型?这个似乎也不能遗传
其他的还有什么?epigenetics的marker? |
|
T*****n 发帖数: 274 | 6 1. 确定是Δ/Δ而非fl/fl?
2. Target gene只有某个exon/UTR被deleted, 而你的real time PCR引物pick up了没有
被敲掉的genomic DNA contaminant。检查引物看它们的genome locus是否隔得够远? |
|
h******y 发帖数: 351 | 7 我的建议是,在和老板争论是否有deletion之前,利用其他的办法证实你的PCR所发现
的结果。PCR是很容易出错的,如果你能利用Southern证实homozygous的deletion确实
存在,相信你的老板是会接受你的结论的。
Constance L. Cepko实验室最近retract了G&D上的一篇文章,很好的说明了PCR的易错
性。有兴趣的可以看看。
http://genesdev.cshlp.org.mutex.gmu.edu/content/25/12/1344.long
“We are writing to clarify the interpretation of the results from our above
-mentioned paper. In this study, we used two methods to examine the
structure of RNA from the mouse Math5 (Atoh7) locus. Our initial
characterization was an analysis of the Math... 阅读全帖 |
|
g***y 发帖数: 201 | 8 1. the mutated allele had dominant negative effect-most likely
2. gene dosage effect-less likely
3. There is a second random mutation happened, which is closely linked with
your targeted locus--very unlikely
possible solution:
1.You may try to breed your chimera with different mouse strains. Sometimes
mutant phenotypes are affected by genetic background.
2. If you are really desperate, you may try ovarian transplantation. |
|
g*********3 发帖数: 177 | 9 各位有沒有點子或者設想?
目前的方法有DNA affinity purification ,標題的那篇文章是09年的,PICH,但是聽說
還沒有實驗室能重複LNA-based PICH.
大家隨便說說。(不知道輸入法怎麼變成了繁體~) |
|
s******s 发帖数: 13035 | 10 看不懂。连什么protein都不知道,怎么western? |
|
s******s 发帖数: 13035 | 11 头疼啊。我有一个project也在试类似的东西,做不出啊 |
|
g*********5 发帖数: 2533 | 12 可以合成这段DNA,生物素标记,然后裂解细胞核,mix overnight
然后 Streptavidin Magnetic Beads
这样标记的DNA被拉下来。复合体也被拉下来
设计好对照。 |
|
n********k 发帖数: 2818 | 13 I believe many have been thinking or trying along the line, me as an example
for at least over 6 years I actually came to this lab thinking I could have
the freedom to pursue this as my boss had the similar idea during my
interview...however aftertalking to some experts in the field, I don't
Think it is that easy although i'll for sure make a major effort on this
once I have a job...
X |
|
s******s 发帖数: 13035 | 14 你这个想的太简单了,最多pull下来一些直接bind的transcription factor。
你这段DNA至少要足够大,带所有上下游信息,然后整合,然后ips,然后做成
老鼠,然后找到你要的细胞,然后pull下来的才是真的 |
|
s******s 发帖数: 13035 | 15 其实这个实验的关键不是动手的部分,关键是ms-spec分辨率不够。
如果ms-spec能够发展个5年10年,我看就很好做了
example
have |
|
l****y 发帖数: 398 | 16 sorry to be negative.
But it seems anything based on affinity purification-MS won't work.
As far as i know, people have tried all sorts of tricks (crosslinking,
minichromatin,targeted cleavage, silac)and failed.
The nonspecific background is ridiculous. It is like every protein is there.
Of course, sometimes a lucky person can pick out one interesting protein by
sharp instinct. but in general, it fails.
if you can make it work as well as CHIP,you get a career. |
|
n********k 发帖数: 2818 | 17 That's what I heard too...The sensitivity or the noise versa background
problem...However, I am still wondering whether there would be any way to
alleviate the problem to the extent that it could provide some meaningful
clues...Or maybe have to go with some kind of microfluid CHIP in conjunction
with MS...I am just thinking aloud...lots of idea on the upstream but no
idea about the real experience or downstream indentification part....It
would be such a tech advance in the entire TF/Epigenetic f... 阅读全帖 |
|
n********k 发帖数: 2818 | 18 Sounds some insider on this...Thanks for the comments...Could you please
elaborate a bit more, what about those very abound TF and known to bind an
element? I am talking about taking a know TF to verify the tech...what's
sensitivity or noise V signal in such scenario...
there.
by |
|
f**u 发帖数: 346 | 19 我觉得这个话题挺有意思的。
难道有很多人试过干这个事情了吗?还是有人试过很多次了?
如果那样的话我挺佩服的,做这个事情需要的细胞数量非常非常大吧。
这个做那么一次的需要多少投入多少时间呀,
如果有人能做三次真得很让人佩服。
there.
by |
|
f**u 发帖数: 346 | 20 要做到和ChIP一样容易好像很难,关键是蛋白质没法扩增。
一般MS的最低要求是0.1 picomole的蛋白,如果你用细胞就需要10^11以上,
也就是几百升的数量级,这个是没法减少的。
当然如果你有办法用动物,一只老鼠应该足够了。
conjunction |
|
n********k 发帖数: 2818 | 21 Thanks I c now...I am very ignorant about the MS part...That's a huge amount
of
cells---to be a bit more accurate:)) it is about 6X10^10...that's about 6000
plates of 10M cells per plates...Provided one could get absolute no-
background...u think we are doomed on this, at least for now?...I am not
gonna giving up
thinking along this, maybe I am being stupid now:))) |
|
y******8 发帖数: 1764 | 22 Design a genetic screening first, then CHIP validation.
Why choose the difficult path?
amount
6000 |
|
f**u 发帖数: 346 | 23 呵呵,你是在假设纯化的产率是100%。
我也没很仔细地想过这个技术只不过这个技术的原理并不复杂,
ChIP出来了那么多年肯定有人想过也做过,
但是目前还没成气候,所以肯定是有一些技术问题在里面的。
当然有技术问题也就说明有机会,并不一定是坏事情。
另外,需要细胞多这个很烦,但是对于去除背景来说也许是好事情。
刚才去看了看楼主提到的文章,他们用telomere的主要原因就是因为拷贝数大,
一个细胞有100个左右,这样细胞用量可以减少两个数量级。
amount
6000 |
|
n********k 发帖数: 2818 | 24 Yes, that's why they worked on telomere...For MS, what's the bottle neck in
term of the sensitivity...how come it needs 10^11 molecules?? I just
searched on line and tried to figure that one out myself...it seems for some
serum protein, it could detect 1pg/ml or even lower...depending on the
loading volume, that could be a lot lower than 0.1pM right?.... |
|
n********k 发帖数: 2818 | 25 Absolutely agree, but it would be a much more direct and economical way(
potentially more powerfully too) if it didn't techniqually seem impossible
as of now... |
|
g*********3 发帖数: 177 | 26 感谢各位大神的帮助。
我倒是觉得MS resolution的问题不是很大,neverthink想的那个点子其实正是我们在
做的,microfluid技术说不定可以部分解决binding的问题。 我个人觉得最大的问题是
如何design DNA oligonucleotide的问题。 It depends on how to define enhancer.
..
当然了,这些都是in vitro的。如果能in vivo就好了。 |
|
g*********3 发帖数: 177 | 27 谢谢!
确实这个问题挺困难,目前还没有什么号的方法。我很头疼,难道我只能靠猜测几个
potential TFs, 然后做ChIP-seq了 (对于我的project)?这样发现某些新东西的可
能性就不大了。各位说的都很有道理:
Shakuras说的DNA长度的问题,目前很多人做的(Bulyk @ Harvard)都只是10-mer,至
于想模拟chromatin之类那么长的DNA我觉得很难。
难道是个悲剧。。。。这个project挺有意思的,以前发过贴问过版上各位可惜没人回
复。
即将拿到ChIP-seq的数据,但是不知道下一步怎么走。如各位说的:
怎样Purify specific的TFs complex,太难了。
conjunction |
|
g*********3 发帖数: 177 | 28 谢谢。我还没做过这个实验。
是nonspecific太多呢,还是基本purify nothing呢?我怎么感觉到in vitro affinity
purification因为不是Physiology condition,可能啥都钓不到。当然,如果有什么
东西,很有可能是nonspecific了。
顺带问下:如果把这个作为phd project来做。。。是不是SB了。
there.
by |
|
f**u 发帖数: 346 | 29 你是从哪个网页上看到这个灵敏度的?
我记得前一阵子听报告人家做protein MS的说要0.1 pmole,
我刚才差了一下网上也差不多。
in
some |
|
f**u 发帖数: 346 | 30 看了半天原来是要坐in vitro,这样就容易很多了。
我记得二十几年前就有人用nuclear extract来泡合成的oligo,
然后用gel shift作为readout来纯化DNA binding protein了。
困难的是in vivo。
enhancer. |
|
f**u 发帖数: 346 | 31 也不用那么麻烦吧,还iPS什么的,直接作个转基因老鼠就可以了。
其实这些上游准备都不算什么,关键是纯化的protocol。 |
|
l****y 发帖数: 398 | 32 In an orbitrap, 100fmol peptide usually give a normalized counts of 10^7 -
10^8. The noise is around 10^4 -10^5. |
|
s******s 发帖数: 13035 | 33 well, 因为最近研究epigenetics,所以思路里面最好的control
就是分化细胞和ips过再重新分化的细胞 |
|
s******s 发帖数: 13035 | 34 大家参考一下。其实我是没有精力做的,有人成功的话给我acknowledge就够了。
找你要研究的基因的BAC,然后再E.Coli里面通过homologous recombination在
有兴趣的片段(当然要比较长的)两头加rare的RE site或者ZFN一类的,随便折
腾;里面再加点strong的aptamer, 或者变态点,加一堆lacO.
这玩意儿做成转基因IPS或者ES,好处是可以多copy, 十个八个没问题吧,更多
应该也行吧? 然后做成转基因小鼠,取你有兴趣的特定组织,或者细胞培养,
做ChIP的前面部分就是fix,洗掉细胞质,然后instead of sonication, 用RE
一顿乱切,配合aptamer purification. 你可以先size select然后affinity,
或者先affinity然后size select, 或者几种不同的方法series affinity,
anyway, 拿到东西decrosslinking上ms spec, 当然前面crosslink的时候不能
用glycine quench.
这个思路欢迎大家补充。 |
|
f**u 发帖数: 346 | 35 握手。我下午也想了想,怎么能够不用DNA杂交来分离目标片断。
基本思路和你想得类似,但是还没你想得那么细。
然后刚刚上网艘文献,发现有人用了跟你说得几乎一样的方法在E.coli里做过了。
DNA sampling: a method for probing protein binding at specific loci
on bacterial chromosomes
看来这个idea还是有不少人想到过的。
做细菌好呀,一升细菌足够用了,如果是小鼠组织的话可能要公斤级。
另外就是我更加确定了肯定有人做过或者正在做类似的事情,但是有技术问题。 |
|
n********k 发帖数: 2818 | 36 tons of people thinking along this way---me and my boss included...Last year
when I finally had some time to fool around:)) and talked to my friend and
asked for his Gal4 etc constructs and he recommended me to use Lac. But he
also essentially point it out that as of now it is extremely hard if not
possible and tons of people are thinking about this problem... |
|
s******s 发帖数: 13035 | 37 这个思路我想过几次了。思路不难,技术上也不难,难的是一个
人要所有的技术都会,还要一直optimize, 说到底生物还是实验科学,
idea不值钱啊 |
|
s******s 发帖数: 13035 | 38 小鼠有些组织细胞很小很多的,比如thymus。 不过这玩意儿貌似细胞分开了
还会重新黏在一起,不知道这么多thymus chip怎么做的,我试了一下不太方便 |
|
s******s 发帖数: 13035 | 39 //nod. 那个PICH的paper我bet至少有100个实验室试过,竟然能让
那个家伙做出来,很佩服
year
and |
|
n********k 发帖数: 2818 | 40 So that would mean the sensitivity/resolution could go lower than 0.1pM,
right, maybe 10-100 fold? That would be doable for some cell culture or
tissue work. BTW, I am an MS-idiot:)))... |
|
n********k 发帖数: 2818 | 41 negative selection, tons of DNA? so very sticky, treat with DNase? just my
pure imagination... |
|
s******s 发帖数: 13035 | 42 说实话,那个PICH我还没有全部理解。
理论上,sonicate以后,不被protein protect的DNA应该断掉了,
但是杂交的时候,被protein crosslink的部分应该不容易杂上吧?
是不是和前面有人说的telomere的所谓特殊结构有关系?
year
and |
|
n********k 发帖数: 2818 | 43 So, what's the bottle-neck for MS sensitivity/resolution? |
|
|
f**u 发帖数: 346 | 45 这个和sonication的具体条件有关吧。
DNA的断点和蛋白结合点之间可能会有几百个bp,足够杂交了。
另外如果探针序列在两个蛋白结合位点之间也有可能杂上,
假设DNA在这两个蛋白结合位点之间没有被打断的话。 |
|
n********k 发帖数: 2818 | 46 but 0.1pM is still a lot---that's 10^11....I believe the DNA Sequencer could
do way better than that....I just don't understand what's the problems? Any
other sensitive approaches than MS available to identify an unknown
molecules? |
|
n********k 发帖数: 2818 | 47 Sounds reasonable to me...Biology is about trying to see whether it will
work...sometimes it just doesn't work the way we can reason... |
|
s******s 发帖数: 13035 | 48 所以我觉得有运气的因素。
做ChIP类得,sonicate成monomer, 基本上两头有个一二十bp的free dna
顶多了吧,不太会有几百。感觉上可能是因为telomere那里可能是一堆
蛋白+DNA团在一起,sonicate不开。如果是这样的话,这个PiCH可能不容
易在其他普通的chromatin上做 |
|
f**u 发帖数: 346 | 49 其实DNA sequencing和protein mass spec对于样品数量在一个数量级上。
一个bp的分子量大概是675,0.1pmole的1kb PCR product大概是70ng,
基本就是你送测序的量了。
could
Any |
|
m******5 发帖数: 1383 | 50 我觉得你和我回帖的思路差不多…… RE 切fix过的,southern + western,都是size
sepcific的
前面有人说还不知道是什么蛋白怎么办,其实我觉得关键是有一个idea这个序列是
activate的还是repressive 的 可以用一些已知的co-activator或者co-repressor来
locate. 要是序列的功能都不知道就难办了…… |
|