n**********g 发帖数: 196 | 1 OK, you want basic chemistry and here is basic chemistry.
“Since ssDNA
can not bind to the editing enzyme alone in vitro, the conclusion will be
the enzyme needs the second substance that is a conformation-ready genomic
DNA to form a functional enzyme-ssDNA-genomic target DNA complex to complete
the reaction. ”
----This is so wrong. There could be 100 reasons why an enzyme does not bind
a substance. For example, temperature, pH. Most importantly, it directly
contradicts Han’s data (Fig 1,2).
“As... 阅读全帖 |
|
n**********g 发帖数: 196 | 2 1) The current low (zero) verified success rate is sufficient for people to
question the story.
When will your statistic data collection be completed? Let's talk about the
success
rate then.
Your comments are so interesting. If you do not know the science, just be
patient and wait for more answers.
2) “Do you think 1.3 mm difference (resolution)
can be achieved on peaks with a width of 1.5-2 mm?” The answer is yes
because of Fig 4e. Your statement is “difference
sample preparation” is completely... 阅读全帖 |
|
n**********g 发帖数: 196 | 3 You comments are funny and clearly show you don't understand gene editing or
this paper at all. The reason you found evidences against Han are weak is
simply because you picked weak ones to fight against.
1) You avoided to answer the lack of reproducibility. You want to "window"
as long as possible to deny this concern. But Han distributed plasmids to
many Chinese laboratories during talks and more than 100 labs acquired his
NgAgo plasmid through Addgene. Yet there is only a single (unverified) ... 阅读全帖 |
|
发帖数: 1 | 4
--你肯定没有做过ssODN用于hES的编辑,做过的话,你应该知道效率有多低,不要太
trust张峰的paper,是可行,但是效率低得吓人。没有足够时间和人力,建议上筛选。
--不太care这种情况,我是收集编辑成功的细胞,效率再低,只要我流式筛选到符合要
求的细胞,就成功。不会像ssODN那样需要很愚蠢的分单细胞去挨个手工筛选,想象一
下让你筛选2000个单克隆的genotyping是哪种工作量,很多实验室筛选到2000个单克隆
还没有正确编辑的细胞。
--相比筛选2000个单克隆,你觉得哪个复杂? 克隆质粒一个星期就可以搞定。 |
|
r*******y 发帖数: 48 | 5 you just need synthesized ssODN as a template for transfection; it is not
necessary to subclone it into plasmd and then release it. |
|
c******d 发帖数: 647 | 6 非行家,不过最近也在做HDR template,
网上关于sgRNA的资源就不说了,但是关于HDR的简直太少
真的是有做的人可能就直接合成ssODN了,可是我想插入比如FP,两边的arm也想1-1.
5kb 长,所以现在也想自己设计plasmid,现在还没做到那步
Zhang的paper说应该linearize donor vector;但我看到别的source说如果用
linearized donor,会增加随机integration的可能,用全plasmid就可以了
cas9 |
|
l***y 发帖数: 638 | 7 HDR插长片段用全plasmid就行,ssODN在人的细胞系里效率很低 |
|
s********r 发帖数: 312 | 8 请教下有经验的同学,看了文章说一般efficiency只有1%,很拼人品的样子,而且如果
挑到了mixed colony也没办法筛选。是不是plasmid donor带selection marker更保险
些, 虽然还要麻烦再挑一次克隆去掉marker。 |
|
|
o**4 发帖数: 35028 | 10 我的donar上没有筛选标志,也想用crispr跟ssODN做knockin,按照你说的效率极低,
heterozygous的效率都只有千分之一,那homozygous是不是没戏了?
要是你会挑多少克隆试试呢?
非常感谢。 |
|
o**4 发帖数: 35028 | 11 请问一下你的donor vector咋设计的呢?是ssODN吗?用了多长的homologous arm呢? |
|
s********r 发帖数: 312 | 12 donor如果用ssODN的话效率比较低吧,筛克隆比较麻烦。 |
|
发帖数: 1 | 13 在人的干细胞里边,除非你使用那个ssODN,这个据很多人反馈是你不筛够2000个克隆
,想都不用想成功。另外能够想到的foot-print free的编辑方法就是结合piggyBac了。
确实不太方便,文章里边报道的加个HSV-DTK,使用FIAU做阴性筛选,百分之60-70.实
际效果是,筛选70-80个克隆,有一个抗FIAU的克隆是成功去除掉了筛选标签。 |
|
发帖数: 1 | 14 我比较希望韩春雨是因为隐藏了某些东西而导致大家无法重复。但是这种可能性很小,
实话实说,自从CRISPR-Cas9的出现就已经大大的降低了基因编辑的门槛,至少从基因
编辑的原理来说,不可能复杂到哪里去。基因编辑的过程就是给细胞提供分子剪刀和修
复模板,你的NgAgo通过载体运到细胞内部表达不可能比Cas9还要困难(如果你的文章
中用的NgAgo是你导入了某些突变的话,你并没有在文章中提及,而故意给大家发放错
误的质粒,我不知道这个大家该怎么看,至少作为一个人正常人都不会这么干的吧)。
另外5‘-P-ssDNA的运输进细胞也不可能会比90bp长的ssODN困难。唯一可以隐瞒的就是
如何降低5‘-P-ssDNA对细胞的毒性。不管怎么看,这篇文章都很weired。从他在北京
大学做报告来看,讲得确实很烂,不像是一个受过正规训练的科研工作者。 |
|
n**********g 发帖数: 196 | 15 It's funny that you just picked the soft accusations to argue against.
Those are not "major concerns", especially PS. PS is intended to explain the
real major concerns, which are
1) lack of reproducibility
2) band shift pattern are conflicting with each other in Fig4a and 4e.
3) the integrated intensity of bands in Fig 4c is not consistent with their
size and molar ratio.
ssODN transfection tricks? No one has issue transfecting it with CAS9.
Take your time and read your book. Then come back with... 阅读全帖 |
|
n**********g 发帖数: 196 | 16 这人的帖子正说明了韩文章的费解之处(甚至说是不合理之处),因为转入质粒后,尤
其是HEK293这种会扩增质粒的cell line,蛋白会持续表达至少数天,没有道理晚转染
ssODN几小时到24小时就不work。
令,如果没读过三遍韩文的就不要发言评论,那么这里我猜会是清一色的质疑文章。 |
|
发帖数: 1 | 17 我倒不那么认为,我不觉得作者会那么low。问题是人家就那么随随便便一测就能发现
如此多的随机突变,这个从概率上来说,已经可以说明问题了。不要因为不喜欢别人的
结果而坚持要推翻别人的结论。
在supplementary的数据中写道:FVB/NJ mice, which are inbred and homozygous
for the retinal degeneration 1 (rd1) allele of Pde6b, were purchased from
The Jackson Laboratory, (Bar Harbor, ME). Co-housed FVB/NJ mice without
CRISPR-mediated correction were used as the functional-deficient control.
Briefly, an sgRNA- expressing plasmid had been coinjected, into FVB/NJ
zygotes, with the single-stranded oligodeoxynucle... 阅读全帖 |
|
发帖数: 1 | 18 CRISPR 的效率和安全方面造假太多, 你只要自己用过你就知道CRISPR技术中那个
donor的随机整合有多么恐怖, 那个ssODN donor效率有多低。
生物信息学压根就是个编故事的学科,除非实验生物领域不存在造假问题,建立在造假
的实验结果上的生物信息相关研究都是扯淡。
immunotherapy |
|