c*b
2 楼
大学里的博士后,最近由于一个很特别的原因老板不得不把我的工资长到接近6万,虽
然是暂时,但想利用这申请EB1-B。可是title还是“senior research fellow”。这个
title能申请EB1-B吗?谢谢!
然是暂时,但想利用这申请EB1-B。可是title还是“senior research fellow”。这个
title能申请EB1-B吗?谢谢!
t*n
3 楼
有哪些决定性优势?
1. MAKE OFFER ASAP?
2 MAKE HIGHER OFFERS THAN LISTING PRICE?
不知道怎么才能抢到SHORT SALE>
谢谢大家.
1. MAKE OFFER ASAP?
2 MAKE HIGHER OFFERS THAN LISTING PRICE?
不知道怎么才能抢到SHORT SALE>
谢谢大家.
y*u
4 楼
如题
做了几批一个transcription factor的CHIP-Seq
出来的peak pattern都和DnaseI很像,是否正常?(detail都像)
做了几批一个transcription factor的CHIP-Seq
出来的peak pattern都和DnaseI很像,是否正常?(detail都像)
a*9
5 楼
没问题啊。没图片一般是你ID输错了吧。
d*t
6 楼
和title关系不大,主要看你的offer letter怎么写
f*n
9 楼
居然记错ID了,看来人不服老不行吧。
h*d
10 楼
直接找那个卖房的经纪人当你的买方经纪人。
往往比人家出价低都能抢的到,还可以要点回扣。
往往比人家出价低都能抢的到,还可以要点回扣。
B*5
13 楼
我约一年前买了一栋short sale房子. 当我的出价买short sale房子的时候, 至少有10
人或以上bid for the house. 通常你要overbid 20K to 30K左右. 幸好我有一个很好
的经纪人. 我买了我的房子slightly lower than the asking price. 同时,银行支付
bank loan closing cost.
人或以上bid for the house. 通常你要overbid 20K to 30K左右. 幸好我有一个很好
的经纪人. 我买了我的房子slightly lower than the asking price. 同时,银行支付
bank loan closing cost.
y*u
14 楼
这批input出问题没拿到
能否简单说一下你说的这种鉴定HS site的方法?
另外你说的这种因为bias导致的peak会很高么?
【在 M********a 的大作中提到】
: 大部分的TF binding site都是在nucleosome free region,也就是DnaseI HS sites,
: 但是如果chromatin没有solublize好,因为crosslink和sonication的bias,会导致
: input也出现DnaseI HS的pattern,甚至有一种方法就是用这个原理去identify HS
: sites的。一般ChIP protocol用150mM的盐,但是最好是300mM,盐浓度太低会导致
: chromatin溶解不了。
能否简单说一下你说的这种鉴定HS site的方法?
另外你说的这种因为bias导致的peak会很高么?
【在 M********a 的大作中提到】
: 大部分的TF binding site都是在nucleosome free region,也就是DnaseI HS sites,
: 但是如果chromatin没有solublize好,因为crosslink和sonication的bias,会导致
: input也出现DnaseI HS的pattern,甚至有一种方法就是用这个原理去identify HS
: sites的。一般ChIP protocol用150mM的盐,但是最好是300mM,盐浓度太低会导致
: chromatin溶解不了。
m*5
17 楼
Agree!
如果peaks都很significant我觉得不用担心。input bias再厉害一般都是出small
peaks.
很多transcription factors binding peaks , 特别是binding sites多,测序深度高
的,都和dnaseI hypersensitivity peaks很相似。
有一些protocol甚至人为增加了input bias来导致available for chip的chromatin 都
在dnasei hypersensitivity area。
降低背景提高灵敏度
【在 M********a 的大作中提到】
: 大部分的TF binding site都是在nucleosome free region,也就是DnaseI HS sites,
: 但是如果chromatin没有solublize好,因为crosslink和sonication的bias,会导致
: input也出现DnaseI HS的pattern,甚至有一种方法就是用这个原理去identify HS
: sites的。一般ChIP protocol用150mM的盐,但是最好是300mM,盐浓度太低会导致
: chromatin溶解不了。
如果peaks都很significant我觉得不用担心。input bias再厉害一般都是出small
peaks.
很多transcription factors binding peaks , 特别是binding sites多,测序深度高
的,都和dnaseI hypersensitivity peaks很相似。
有一些protocol甚至人为增加了input bias来导致available for chip的chromatin 都
在dnasei hypersensitivity area。
降低背景提高灵敏度
【在 M********a 的大作中提到】
: 大部分的TF binding site都是在nucleosome free region,也就是DnaseI HS sites,
: 但是如果chromatin没有solublize好,因为crosslink和sonication的bias,会导致
: input也出现DnaseI HS的pattern,甚至有一种方法就是用这个原理去identify HS
: sites的。一般ChIP protocol用150mM的盐,但是最好是300mM,盐浓度太低会导致
: chromatin溶解不了。
M*a
20 楼
interesting.我才知道有这样做的,这样的话control就更显重要了。从USCS browser
上看Encode的有些IgM/G control, peak还是很pronounced
【在 m******5 的大作中提到】
: Agree!
: 如果peaks都很significant我觉得不用担心。input bias再厉害一般都是出small
: peaks.
: 很多transcription factors binding peaks , 特别是binding sites多,测序深度高
: 的,都和dnaseI hypersensitivity peaks很相似。
: 有一些protocol甚至人为增加了input bias来导致available for chip的chromatin 都
: 在dnasei hypersensitivity area。
: 降低背景提高灵敏度
M*a
22 楼
请问能给我些reference吗?我想看看他们这样做的话最后是如何normalize的,或者根
本就是用raw reads就说明问题了?
我本身不是做TF的,这方面paper读得不多,刚才随便查了一下发现好多好多paper图上
竟然都是直接展示 number of reads.这太不可思议了,竟然没有人站出来质疑?
【在 m******5 的大作中提到】
: Agree!
: 如果peaks都很significant我觉得不用担心。input bias再厉害一般都是出small
: peaks.
: 很多transcription factors binding peaks , 特别是binding sites多,测序深度高
: 的,都和dnaseI hypersensitivity peaks很相似。
: 有一些protocol甚至人为增加了input bias来导致available for chip的chromatin 都
: 在dnasei hypersensitivity area。
: 降低背景提高灵敏度
本就是用raw reads就说明问题了?
我本身不是做TF的,这方面paper读得不多,刚才随便查了一下发现好多好多paper图上
竟然都是直接展示 number of reads.这太不可思议了,竟然没有人站出来质疑?
【在 m******5 的大作中提到】
: Agree!
: 如果peaks都很significant我觉得不用担心。input bias再厉害一般都是出small
: peaks.
: 很多transcription factors binding peaks , 特别是binding sites多,测序深度高
: 的,都和dnaseI hypersensitivity peaks很相似。
: 有一些protocol甚至人为增加了input bias来导致available for chip的chromatin 都
: 在dnasei hypersensitivity area。
: 降低背景提高灵敏度
m*5
24 楼
Haha
What kind of sequencing do you usually do? You really know a lot! I also
learned something from what you just pasted here.
What I thought Is that you can always use local background to normalize as a
means to determine how significant your peak is compared to local
background.
For the pictures you just attached,I bet those peaks that are present in
input must have been included in the encode blacklist( document those
regions that would easily come up due to repetitive sequence and library
construction)
Meanwhile, what is the scale bar of the picture you just attached? From the
Y axis you showed I don't think those peaks are actually pronounced. And the
peak that are shown in IgG/M are also shown in input, suggest it is just
some kind of bias during library construction
【在 M********a 的大作中提到】
: 请问能给我些reference吗?我想看看他们这样做的话最后是如何normalize的,或者根
: 本就是用raw reads就说明问题了?
: 我本身不是做TF的,这方面paper读得不多,刚才随便查了一下发现好多好多paper图上
: 竟然都是直接展示 number of reads.这太不可思议了,竟然没有人站出来质疑?
What kind of sequencing do you usually do? You really know a lot! I also
learned something from what you just pasted here.
What I thought Is that you can always use local background to normalize as a
means to determine how significant your peak is compared to local
background.
For the pictures you just attached,I bet those peaks that are present in
input must have been included in the encode blacklist( document those
regions that would easily come up due to repetitive sequence and library
construction)
Meanwhile, what is the scale bar of the picture you just attached? From the
Y axis you showed I don't think those peaks are actually pronounced. And the
peak that are shown in IgG/M are also shown in input, suggest it is just
some kind of bias during library construction
【在 M********a 的大作中提到】
: 请问能给我些reference吗?我想看看他们这样做的话最后是如何normalize的,或者根
: 本就是用raw reads就说明问题了?
: 我本身不是做TF的,这方面paper读得不多,刚才随便查了一下发现好多好多paper图上
: 竟然都是直接展示 number of reads.这太不可思议了,竟然没有人站出来质疑?
m*5
25 楼
Meanwhile what you showed are almost all input tracks. From what I know
Encode seldomly use non specific IgG/igM for sequencing after carefully
characterization
of the antibody. (Or maybe they even do so during early days)
【在 M********a 的大作中提到】
: 我随便搜了几个control,不同cell type的,对比IMR90的DNaseI HS (top),有一个
: peak在这几个control里都非常强,你总不能拿来当hit吧。
Encode seldomly use non specific IgG/igM for sequencing after carefully
characterization
of the antibody. (Or maybe they even do so during early days)
【在 M********a 的大作中提到】
: 我随便搜了几个control,不同cell type的,对比IMR90的DNaseI HS (top),有一个
: peak在这几个control里都非常强,你总不能拿来当hit吧。
M*a
26 楼
谢谢你的回答,我也学习了:)
整个window 200k, 这个window一般y值多大算significant?ChIP-qPCR至少也有对
input和mock IgG的control, 为什么ChIP-seq就只需要看local background? 所有做
ChIP-seq的都了解Blacklist吗?他们都会把这些hit自动去掉吗?那还有一些cell type
specific的input peak 没有在blacklist里的怎么办?
a
【在 m******5 的大作中提到】
: Haha
: What kind of sequencing do you usually do? You really know a lot! I also
: learned something from what you just pasted here.
: What I thought Is that you can always use local background to normalize as a
: means to determine how significant your peak is compared to local
: background.
: For the pictures you just attached,I bet those peaks that are present in
: input must have been included in the encode blacklist( document those
: regions that would easily come up due to repetitive sequence and library
: construction)
整个window 200k, 这个window一般y值多大算significant?ChIP-qPCR至少也有对
input和mock IgG的control, 为什么ChIP-seq就只需要看local background? 所有做
ChIP-seq的都了解Blacklist吗?他们都会把这些hit自动去掉吗?那还有一些cell type
specific的input peak 没有在blacklist里的怎么办?
a
【在 m******5 的大作中提到】
: Haha
: What kind of sequencing do you usually do? You really know a lot! I also
: learned something from what you just pasted here.
: What I thought Is that you can always use local background to normalize as a
: means to determine how significant your peak is compared to local
: background.
: For the pictures you just attached,I bet those peaks that are present in
: input must have been included in the encode blacklist( document those
: regions that would easily come up due to repetitive sequence and library
: construction)
M*a
27 楼
我也不明白input-igG是什么意思,input normalized by igG?
我看到有光是写input的,那个估计就是raw input吧?
【在 m******5 的大作中提到】
: Meanwhile what you showed are almost all input tracks. From what I know
: Encode seldomly use non specific IgG/igM for sequencing after carefully
: characterization
: of the antibody. (Or maybe they even do so during early days)
我看到有光是写input的,那个估计就是raw input吧?
【在 m******5 的大作中提到】
: Meanwhile what you showed are almost all input tracks. From what I know
: Encode seldomly use non specific IgG/igM for sequencing after carefully
: characterization
: of the antibody. (Or maybe they even do so during early days)
m*5
28 楼
I dont know enough to answer how significant is significant. I have seen
people making point on peaks that are really small( in that case, input
tracks are also show). It also has something to do with how 'peaky the peak
is. Mainly the shape of the peak.
For the second question: for one, in chip q pcr, it is hard to see the
pattern, so counting relative enrichment is simply the only thing they cound
do out of the situation.
I personally don't trust chip q pcr because some antibody simply have strong
global binding compared to igG sample.
type
【在 M********a 的大作中提到】
: 谢谢你的回答,我也学习了:)
: 整个window 200k, 这个window一般y值多大算significant?ChIP-qPCR至少也有对
: input和mock IgG的control, 为什么ChIP-seq就只需要看local background? 所有做
: ChIP-seq的都了解Blacklist吗?他们都会把这些hit自动去掉吗?那还有一些cell type
: specific的input peak 没有在blacklist里的怎么办?
:
: a
people making point on peaks that are really small( in that case, input
tracks are also show). It also has something to do with how 'peaky the peak
is. Mainly the shape of the peak.
For the second question: for one, in chip q pcr, it is hard to see the
pattern, so counting relative enrichment is simply the only thing they cound
do out of the situation.
I personally don't trust chip q pcr because some antibody simply have strong
global binding compared to igG sample.
type
【在 M********a 的大作中提到】
: 谢谢你的回答,我也学习了:)
: 整个window 200k, 这个window一般y值多大算significant?ChIP-qPCR至少也有对
: input和mock IgG的control, 为什么ChIP-seq就只需要看local background? 所有做
: ChIP-seq的都了解Blacklist吗?他们都会把这些hit自动去掉吗?那还有一些cell type
: specific的input peak 没有在blacklist里的怎么办?
:
: a
M*a
29 楼
make sense
我觉得encode如果有标准的workflow,input质量还不错,加上去掉blacklist,那不
normalize也就算了,但是如果一般的lab,尤其是没经验的,意识不到crosslinking 和
sonication对input的影响,如果没做好,弄出一堆bias,又不normalize,岂不是结果都
不可靠了。。我没有仔细研究过,但愿如你所说,bias导致的peak值都比较小。我见过
normalize之后log2的值是0.4的,作者还claim enrichment。。。
peak
cound
strong
【在 m******5 的大作中提到】
: I dont know enough to answer how significant is significant. I have seen
: people making point on peaks that are really small( in that case, input
: tracks are also show). It also has something to do with how 'peaky the peak
: is. Mainly the shape of the peak.
: For the second question: for one, in chip q pcr, it is hard to see the
: pattern, so counting relative enrichment is simply the only thing they cound
: do out of the situation.
: I personally don't trust chip q pcr because some antibody simply have strong
: global binding compared to igG sample.
:
我觉得encode如果有标准的workflow,input质量还不错,加上去掉blacklist,那不
normalize也就算了,但是如果一般的lab,尤其是没经验的,意识不到crosslinking 和
sonication对input的影响,如果没做好,弄出一堆bias,又不normalize,岂不是结果都
不可靠了。。我没有仔细研究过,但愿如你所说,bias导致的peak值都比较小。我见过
normalize之后log2的值是0.4的,作者还claim enrichment。。。
peak
cound
strong
【在 m******5 的大作中提到】
: I dont know enough to answer how significant is significant. I have seen
: people making point on peaks that are really small( in that case, input
: tracks are also show). It also has something to do with how 'peaky the peak
: is. Mainly the shape of the peak.
: For the second question: for one, in chip q pcr, it is hard to see the
: pattern, so counting relative enrichment is simply the only thing they cound
: do out of the situation.
: I personally don't trust chip q pcr because some antibody simply have strong
: global binding compared to igG sample.
:
y*u
30 楼
非常感谢楼上各位的comment!我就多测一些好了
m*5
31 楼
I too am actually a newbie so would greatly appreciate someone with more
advanced knowledge to join this topic.
My understanding is that once you are convinced that your antibody works
well, such as it can pull down the protein of your interest after extensive
crosslinking, it became easier for you to trust the pattern you obtained in
your sample.
And if you ever did, what is the amount of the target TF you can pull down
in the tissue of your interest? is it very abundant?
For most companies ChIP sequencing project takes a while(few weeks to month)
, so maybe you can try make the best out of your current project while
waiting for sequencing result of your input.
【在 y*********u 的大作中提到】
: 非常感谢楼上各位的comment!我就多测一些好了
advanced knowledge to join this topic.
My understanding is that once you are convinced that your antibody works
well, such as it can pull down the protein of your interest after extensive
crosslinking, it became easier for you to trust the pattern you obtained in
your sample.
And if you ever did, what is the amount of the target TF you can pull down
in the tissue of your interest? is it very abundant?
For most companies ChIP sequencing project takes a while(few weeks to month)
, so maybe you can try make the best out of your current project while
waiting for sequencing result of your input.
【在 y*********u 的大作中提到】
: 非常感谢楼上各位的comment!我就多测一些好了
m*5
32 楼
I am just reading the sono-seq paper you refereed to
http://www.pnas.org/content/106/35/14926.full.pdf+html?with-ds=
How come a paper describing an artifact could be published on PNAS…………
anyhow, it appears that this paper actually taught us to ignore the bias in
input because they found such bias irrelevant to the baseline level of
binding in IgG (as those regions enriched in Input were not enriched in the
IgG control )
false
【在 M********a 的大作中提到】
: FAIRE,还有一个sono-seq也很类似。
: 我觉得如果你没有input的话就比较悬,至少igG control得有,不然的话哪些是false
: positive很难说清楚,到时候reviewer估计也会问。
: 因为每次试验的条件会有些差别,比如reverse crosslink的时间,所以input的
: variability其实差别还是挺大的。Control真的很重要。
http://www.pnas.org/content/106/35/14926.full.pdf+html?with-ds=
How come a paper describing an artifact could be published on PNAS…………
anyhow, it appears that this paper actually taught us to ignore the bias in
input because they found such bias irrelevant to the baseline level of
binding in IgG (as those regions enriched in Input were not enriched in the
IgG control )
false
【在 M********a 的大作中提到】
: FAIRE,还有一个sono-seq也很类似。
: 我觉得如果你没有input的话就比较悬,至少igG control得有,不然的话哪些是false
: positive很难说清楚,到时候reviewer估计也会问。
: 因为每次试验的条件会有些差别,比如reverse crosslink的时间,所以input的
: variability其实差别还是挺大的。Control真的很重要。
y*u
33 楼
顶上去期待更多讨论
相关阅读
对比颜宁和屠呦呦看非诺奖和诺奖科学家的差别在哪里?你们都错了,颜老师出国是因为国内找不到好司机成立个 Nieng Yan 俱乐部好了bioinformatician 职位 @UMN (PhD required)面对明目张胆作假的老板,该怎么办?不同数据库之间ID的转换有在纽约附近做基因检测分析相关工作的吗?Nature communications上的文章够烂急!!访学1年因自费压力大且不是同一个领域,每天很郁闷,想换实验室,求内推微生物,生物信息,生物技术,分子生物学。真正的科技都是掌握在幕后英雄手中的Muse cell 和 iPSC 之争Re: 中科院理化所杭州研究院是什么性质的? (转载)写在转行前谁给预测一下今年Nature, Cell, Science和子刊的影响因子??有谁知道怎么申请失业救济金? (转载)Re: 感觉老师有这么几种档次 (转载)求一篇paper, 有点急。给个原创的,临床相关的aging生物信息研究的例子NIH 重要消息 (转载)施一公在美国有没有place 人啊?