O******e
发帖数: 4845 | 1 俺觉得他还是比较牛，但不是大牛，^_^。
至于智商超不超群，the phenotype of working 20hr/day让人有些犹豫。 |
|
f******u
发帖数: 9 | 2 GENERAL SUMMARY
A postdoctoral position is available immediately in the laboratory of Dr.
Christopher Bartlett in the Battelle Center for Mathematical Medicine to
perform neurogenetic studies correlating genetic variation to gene
expression phenotypes from rodent and human brain tissue samples. Essential
methodologies include using multiplex PCR, RT-PCR, quantitative PCR, tissue
histology, in situ hybridization and immunohistochemistry methods interfaced
microscopy data collection. Experience wi |
|
H*g
发帖数: 2333 | 3 Why you mentioned there should be pretty well overexpression in plant? You
observed a phenotype, or due to antibiotic tolerance? |
|
H*g
发帖数: 2333 | 4 yeah, I feel it is better the author answer the question: why he/she feel
there must be high expression, based on ox phenotype, based on high antibioc
tolerance or based on RT-PCR data? |
|
a********k
发帖数: 2273 | 5 我像是那么高级的人么。。。是说其实幼鱼的phenotype绝大部分都表现为严重的发育
defects，和人类疾病差太远，而成鱼做疾病模型挺好但是很不好做imaging。 |
|
p*****m
发帖数: 7030 | 6 I see，hoho 不过你说的是mutant phenotype吧 我说的更倾向于后天诱发出来的
比如给鱼猛吃脂肪。。 |
|
a********k
发帖数: 2273 | 7 我就是说后天诱发的都要等到至少Juvenile，到时候鱼不咋透明了，而且自发荧光害死
人。
Mutant phenotype大多不适合做疾病模型去晒选。 |
|
X******n
发帖数: 914 | 8 Good point。
看了那两篇文章，rapamycin是我最喜欢的化合物之一。没想到Huang’lab做了这么好
的工作， 不亏是Shreiber的弟子。早听说Mike筛选了无数化合物，手上有上千化合物
有好的phenotype，看来到出东西的时候了。酵母这个模型做筛选还是有优势。
, |
|
b******d
发帖数: 149 | 9 I think it's a descent paper. small molecule screen + hits with sexy
phenotype, but lacks of target and mechanism. |
|
m*********7
发帖数: 5207 | 10 现在刚被接受的这片文章，是2008年12月完成的第一稿。
老板说我动物实验的结果来源于同一批纯化的蛋白，要求用另一个batch重复。
麻烦来了，老鼠不配合，连阳性对照都没有phenotype.
要么就是什么都没注射就死光光。
整个2009年就在折腾这个实验，同时作另一个课题，同时我娃在我肚子里一天一天长大。
2009年年底，终于得到重现性很好的结果，样本也足够大。
2010年二月，我预产期前一天，文章投稿。两天之后我儿子出生。
第一次审稿耗时八周多，在我产假结束之后的一天，得到major revision的决定。
从四月中到五月中马不停蹄的补试验，同时夜里照顾娃，累得快瘫过去。
五月中旬文章re-submission，第二次审稿耗时七周，昨天终于被接受。
此时我娃已经快五个月了。
一桩心事总算放下了。
感慨一下！ |
|
g**********y
发帖数: 423 | 11 是这篇：A Genomic Strategy to Elucidate Modules of Oncogenic Pathway
Signaling Networks
另外，我总觉得他们用binreg(svd+probit reg)来预测pathway activity有个问题，就
是他们选的gene signature based on the correlation coeff between phenotype
and the gene exp。如果你仔细看看Bild 2006 的Nature paper，可以看出每个
pathway选出的gene list数目都在变，有时候75，有时候200，300等等，然后再作svd(
效果等同PCA)。这样的feature selection 到底对不对？
就像我们做clustering，只选top highly diff expressed genes做clustering可能有
问题，也可能没事？ |
|
h********n
发帖数: 4079 | 12 我有三个转基因老鼠: A(+/-), B(+/-), C(+/+) & C (+/-)
其中A(+/+)不能活, B只需要(+/-)
A和B老鼠都是B6, C是FVB
cross breeding 步骤:
1. A(+/-) X B(+/-) --> A(+/-)B(+/-) in B6
2. A(+/-)B(+/-) X C(+/+) --> A(+/-)B(+/-)C(+/-) in B6/FVB
3. A(+/-)B(+/-)C(+/-) X C(+/-) -->
(1) A(+/-)B(+/-)C(+/+)
(2) A(+/-)B(+/-)C(+/-)
(3) A(+/-)B(+/-)C(-/-)
最终要比较(1)(2)(3)的phenotype, 但是由于A&B是B6, C是FVB, 最终三组会出现背景
不一样的情况, 比较结果可能不可靠. 有人告诉我只要数量够多应该可以.
不知道我说清楚了没有,
各位有什么建议吗?
谢谢. |
|
D*a
发帖数: 6830 | 13 a++为什么不能活？knockin么？是的话不能用－/－，应该是＋/0
c++也是一样的问题
background你要是没有办法弄成一样的就照直写就行了，也可以说这样有variation所
以跟自然生物界的更接近blablabla，
如果你phenotype强的话肯定没问题。另外也看你怎么定义多用几只老鼠了，我们一样
的background用的老鼠都是10只以上，8只就是没办法了，四五只就是实在没办法了 |
|
h********n
发帖数: 4079 | 14 A是transgene, homozygous的A++老鼠发育有问题. 我写成A+/-, 不知道合适不合适.
最终的mating 是 FVB/B6 X B6, 好像后代的background不太一样. 我打算每组用20只
老鼠. 当然phenotype强才好.
有没有啥办法知道FVB/B6 X B6后代的背景情况, 比如测一些gene?
谢谢.
所以跟自然生物界的更接近blablabla，
样的background用的老鼠都是10只以上，8只就是没办法了，四五只就是实在没办法了 |
|
a***e
发帖数: 1010 | 15 (1) for any model organism, you can try homologous recombination to mutate
anything you want.
(2) for model organism like yeast, worm, fly, the best part of them is
genetics. You can do "unbiased" genetic screen to identity anything change
a defined phenotype. |
|
j**e
发帖数: 582 | 16 我到现在也没有搞懂Autism的动物模型的PHENOTYPE性状试验基于的科学道理是什么。
老鼠真的能有Autism吗？ |
|
p*****m
发帖数: 7030 | 17 谢～～回答你刚才的问题 我想要的model是用来快速screen drug candidate的 不是用
来更进一步看效果/safety的（后者也许大鼠和rabbit的model可以接着用）。
之所以想弄个model出来，一个是中风的分子机理还是太不清楚 除了你说的这些以外似
乎没有什么明显的蛋白可以用来做target，和中风specifically相关的好像就没有。。
所以HTS就不好弄，我想如果有个phenotype-based assay（比如说 鱼血管里面弄有机分
子 一照光就aggregate形成血栓 然后可以马上看效果，比如说临近血管堵塞啊 蛋白酶
活性上升啊，细胞necrosis啊，等等，然后就上HTS） 这样可以绕开分子机理未知的问
题。当然这样的model也就只能用来screen，之后的细致分析还需要更精致的模型，比如
血管打结或者弄个针管进去之类的。。。
你觉得这想法靠谱么？
掉。 |
|
w***a
发帖数: 4361 | 18 或者同时设计多个siRNA target同一基因，然后看你感兴趣的phenotype是否sequence
specific。
对于siRNA off target effect，大部分打酱油的paper都直接无视。俺建议你也别费事
算鸟。 |
|
y****2
发帖数: 65 | 19 我的理解
我们知道定位QTL是不准确的。证明一个QTL存在或是不存在是困难的。
在一个不是非常巨大的群体里估计出来的QTl效应是有偏差的。
为啥不放弃QTL的模型，而是仅仅将所有的markers放在一个模型中，来给每个genotype
一个值的估计？--把所有的marker（SNP在这里）同时放在一个模型中，预测每个
marker allele的对于疾病的效应。所有marker alleles的值总和就是这个genotype的
效应。
总之，我们可能不知道QTL是啥，但是我们可以更加准确地预测每个genotype的值。黑
箱操作。需要对于每个群体都单独分析，因为LD不consistent across populations.LD
啊LD,population structure，genetic architecture这些都是association analysis
极端重要的问题。
两个前提，good phenotyping； 第二 if marker –target gene LD is consistent
across populations. |
|
y**s
发帖数: 6809 | 20 弱弱的问啊，能达到那个计算能力吗？
你最后算不出来还是得假定sparse matrix啊
这不就又回到qtl了
我的理解
我们知道定位QTL是不准确的。证明一个QTL存在或是不存在是困难的。
在一个不是非常巨大的群体里估计出来的QTl效应是有偏差的。
为啥不放弃QTL的模型，而是仅仅将所有的markers放在一个模型中，来给每个genotype
一个值的估计？--把所有的marker（SNP在这里）同时放在一个模型中，预测每个
marker allele的对于疾病的效应。所有marker alleles的值总和就是这个genotype的
效应。
总之，我们可能不知道QTL是啥，但是我们可以更加准确地预测每个genotype的值。黑
箱操作。需要对于每个群体都单独分析，因为LD不consistent across populations.LD
啊LD,population structure，genetic architecture这些都是association analysis
极端重要的问题。
两个前提，good phenotyping； 第二 if marker –target |
|
r*****l
发帖数: 457 | 21 你指的单一病因的同种疾病是指
一个phenotype是由多个small effect的基因引起的。只要sample数目够大，理论上都
可以detect到
种病
make |
|
l******u
发帖数: 936 | 22 EMT本身不是新东西, 在动物胚胎学领域已经被研究了好几十年了, 现在只是把这个
phenotype 的概念引入到 epithelial tumor 这个领域. 总的来说属于这个东西
属于tumor microenviroment interaction 的范畴, 这个东西有没有突破不好
说啊, 俺没有这个前瞻性. 反正Bob 老先生是review 和重量级的article一片接
一片的发啊.
Cancer Stem Cell 在液态瘤里做的不错啊, 还比较convincing. 其实Cancer Stem
的概念其实也是1937年就有了, 近十年又炒起来了. |
|
w***a
发帖数: 4361 | 23 Good to know that, hehe
I guess the common practice is to use cocktail to get stable knockdown
and phenotype and then try to identify the strong individual siRNA from
that pool which gives similar silencing and very low off-target as well. |
|
n********k
发帖数: 2818 | 24 As far as I know, Cre has demonstrated effects in lots of line...while in
general loxP effects shall be controlled easily...if it does, u shall have
hypomoph, right? I don't think ur boss's argument for this particular is
that sound...anyway, people use the other breeding strategy (without control
for cre) all the time. If u have a cell death phenotype, then u need to be
extra careful about cre |
|
c********o
发帖数: 135 | 25 即使是一个copy的Cre转基因也应该有control。记得nature发过一篇关于Cre的
toxicity的paper，大致是因为很多转基因的Cre都是很强的promoter驱动的，会影响插
入位点附近的基因表达。所以floxed/floxed并不是floxed/floxed;Cre的最好control
，应该是floxed/wt,Cre。
现在有一些cre是knockin在promoter下游的，至少位置是清楚的。另外一个办法是
knockin在Rosa26的位置。
Homo Cre有phenotype是挺有意思的，原来隔壁实验室就有过，但花了很多时间想搞清
楚插入位点，但最后是不了了之了。 |
|
w***r
发帖数: 709 | 26 是啊，所以让你举个例子。
比如说一个65岁得Parkinson的人，用iPS方法拿到他的neuron cell。我不认为可以在
cell culture 里观察到明显的phenotype。否则，病人早就应该发病了。 |
|
p*****m
发帖数: 7030 | 27 定向诱导和我说的条件不矛盾啊 你要做in vitro drug screen肯定得先能定向诱导了
事实上至少诱导个neuron问题不会太大 中胚层内胚层的组织也许会困难点。
等诱导好了 所谓的摸条件建assay的难度不会高于现在五花八门的in vitro assay
说到这个screen出来的东西到底有多有用 这个就没办法了 事实上所有phenotype-
based screen都是一样的问题 解决方案是geneticists' belief,呵呵
问。 |
|
a****o
发帖数: 1786 | 28 then you should look for synhetic effect phenotype. |
|
A******y
发帖数: 2041 | 29 I am not really in the field. However, there are tons of evidence showing
the amyloid plaques are either protective or doesn't really cause AD. First
, transgenic mice with tons of plaques don't have any disease phenotype.
Second, drugs destroy plaques in human don't do anything to improve the
outcome for patients. Third, scientists have been looking only at one side
of the amyloid, why happen to the cytosol part that's also cleaved off.
Anyone still fighting for amyloid theory to be even pa |
|
b******d
发帖数: 149 | 30
That's exactly the problem that these biotech companies have. They find
some compounds with some phenotype, then quickly go through clinical
trial without truly understanding the target(s) and mechanism. The end
result is endless lawsuits, wasted money in clinic trials and recalls.
In this case, the compound may inhibit the target they desire, but there
are other potential target(s) that they might fail to find/study.
That's why target identification methods are really the bottleneck and
in |
|
D*a
发帖数: 6830 | 31 control说起来可以到四五个组，不过你光control就别想做实验了
我建议你看看你的cre都有谁用，有没有副作用，没有就算了，人家能发一个对照的，
你也能发。另外自己稍微留心下你主要的phenotype是不是cre的作用，留点data， 你
到时候引一下别人的，再data not shown一下就行了。
就算你最后要发cns，也别还没思路的时候第一个实验就上一堆对照，你先有点思路了
，再想想怎么加对照也不迟。 |
|
D*a
发帖数: 6830 | 32 曾经有人跟我老板要老鼠，我老板问他，你的cre能管用么？能证明一下不？然后那人
就不回信了
好几年之后（最近）发了一篇文章，用的另一个人的老鼠，phenotype很莫名，四五分
左右，他还不是第一也不是通讯 |
|
O******e
发帖数: 4845 | 33 Off-target effects might have contributed to what you got. I would try
several
more siRNA oligos targeting different regions of your gene to confirm the
death phenotype. At the same time, try to rescue with resistant constructs.
siRNA |
|
n****n
发帖数: 165 | 34 Rescue assay is really important for your phenotype. I guess your gene maybe
have some function in Mitosis. |
|
i***R
发帖数: 663 | 35 You are the expert ! Talk more about this topic pls. :)
Alexandra C. McPherron in UTSW first got DN ActRIIB mouse line years
ago. The phenotype are striking, very muscular mouse indeed.
http://www.pnas.org/content/98/16/9306.full
Maybe this mouse will also resist to AIDS, cancer etc.
sunny:)))
life
status and
would |
|
a******g
发帖数: 129 | 36 You are coaxed by the sexy title.
Our lab reviewed this paper for Cell. Actually, the only sexy part is tile
and so called EE phenotype. The other pates
are just shit. And that,s why all three reviewers including us rejected them
directly. However,they made a deal with
editor. |
|
m*****n
发帖数: 760 | 37 if this works, that would be exactly what I want.
in this way, I can determine if the phenotypes
are induced specificly by GFP-GOI transgene. |
|
m*********7
发帖数: 5207 | 38 In theory it should work, but then you have to make sure it is not due to
off-target effect, and later rescue the phenotype by making RNAi resistant
GFP-GOI to demonstrate specificity. What's more, what if the effect of
endogenous and ectopic GOI are not additive or synergistic? |
|
G******O
发帖数: 189 | 39 Homo和hetero 在功能上应该是差不多的,但是通常我们用Hetero的原因除了能得到Wild
-Type 对照外,一个重要原因是减少了出现假阳性的PHENOTYPE的几率.因为一般Cre是转
基因进去的，不确定Cre的具体染色体定位，假如碰巧和某基因在相同的位点，HOMO会
带来无关性状。
这度【 在 htscorpion (snoopy) 的大作中提到: 】 |
|
j**********k
发帖数: 296 | 40 按说data quality差些无所谓，但这篇真算是极品了。editor该被fire掉，天雷滚滚的谬误竟然也可以送出去给reviewers？reviewers也是吃屎了，即使print out都该能看出一堆问题。
RETRACTED: VMA21 Deficiency Causes an Autophagic Myopathy by Compromising V-
ATPase Activity and Lysosomal Acidification
Reason: Our paper reported the identification of mutations in the
gene VMA21 in patients with X-linked Myopathy with Excessive Autophagy (XMEA
) and characterized the molecular mechanisms underlying the disease
phenotype. Many of the figure panels in the paper summ |
|
R****n
发帖数: 708 | 41 syber green原理是用allele specific PCR吧，一个primer 3'端用来区分phenotype.
我没做过不太清楚效果。还有什么和ligation结合的，HRM的。你说的太笼统了。
Taqman是用probe,两边一个dye一个quencher，不同的sequence dye不一样，一管PCR搞
定。这个基本上是业界的标准方法 |
|
b******d
发帖数: 149 | 42
known
to
vomiting.
trials
Chi
growth
Good research in a very exciting field: studying natural products and
small molecules, chemical biology and proteomics.
This type of research is kind of opposite to the traditional biomedical
research. It usually starts with some natural products or small
molecules with very interesting known phenotypes, but very little
knowledge of mechanisms and targets. The ultimate research goal is to
find mechanisms and targets. No matter what mechanisms and targets |
|
k*****o
发帖数: 1486 | 43 。。这个，咋说阿。
Pathway比Mechanism要抽象，pathway就是中文“通路，路径”的意思，Mechanism是“
机制”。没有比中文更具体详细的解释方法了。
只要是通路就可以叫pathway, 范围很广，e.g 如上signal transduction的pathway
，metabolic/biosynthetic pathway等等。
Mechanism是机制，比如你看到啥现象了，像是啥化学反应或者一些奇怪的phenotype，
你要解释其背后的原理。
ps个人理解，欢迎排砖。。 |
|
S**********e
发帖数: 1789 | 44 看了这个教授的主页，竟然公布了自己review的文章，题目，杂志名称都有，是不是违反职业道德呢？
Editorial Activities
Journal Peer Reviewer
2009 International Journal of Cancer. (Article) E-cadgerin
transcriptional down-regulation by epigenetic and microRNA-200 family
alterations is related to mesenchymal and drug-resistant phenotypes in human
breast cancer cells, IJC-09-1835.
2009 Molecular Cancer Therapeutics. (Article) Role of reactive oxygen
species in the ability of H-Ras to enhance cell death induced by histone
deacetylase inhibit... 阅读全帖 |
|
r*****l
发帖数: 457 | 45 LD is the non random association between two loci. If one site is your marker and
the other is the causal variance and these two sites are in high LD. Then
you can declare this marker is associated with the phenotype.
hopefully |
|
s******s
发帖数: 13035 | 46 也有可能只是phenotype比较像一类的，具体分析啦 |
|
s*****0
发帖数: 357 | 47 你把两个loci之间的non-random association和genotype-phenotype association的“
association"搞混了。前者是通常所说的linkage，后者才是GWAS常常提到的disease或
者drug response association。
位点如果in strong HWD，如果原因不明，通常会被dropped掉，因为结果不可靠. |
|
p******d
发帖数: 3737 | 48 Several different concepts are being discussed here:
1. The origin of tumor/cancer(single cell or multiple cells)
2. The existence of CSC. Even if the tumor origin is only one cell (
differentiated or tissue-specific stem cell), the whole tumor is
phenotypically heterogeneous, which has been proven by many many
experiments
. Then do all these cells have the same potential to initiate a new tumor?
Or to be more practical, do all these cells have the same potential to met
or relapse after chemo? |
|
h**********r
发帖数: 671 | 49 I don't know anything about this strain. Is it a high GC content actinomyces
? The plasmid perhaps work for the strain.
For the genes involved in TCA cycle, it really depends on the enzymes.
Sometimes there are alternative pathways and metabolic flexibility. There
might also be other isoenzymes. You may add some substance to the media or
clone the similar genes with the same function from other bacteria to
restore the phenotype. |
|
a****d
发帖数: 1919 | 50 depends on how fay you can go with the mechanism to further address the
phenotypes, I would say somewhere from J of Neuroscience, Gene&development
to neuron. |
|
