b********n
发帖数: 673 | 1 不是大拿，回答一下：
30 courses
2. 你认为哪些课程尤其重要
stat inference, multivariate,datamining, measure theory,linear regression,
stochastic process, stochastic differential equation.
comparis
MLE,shrinkage method(like lasso, ridge etc),SVM, linear regression and
generalized linear model
R/SAS/Minitab/Stata/Matlab |
|
A******y
发帖数: 2041 | 2 I thought the ability for western blot to tell the difference has to do with
the linearity of the signal and maximum signal to noise ratio. If you use
film, you are limited by the film's linearity range which I "heard" is
between 2-10, so you need at least two fold differences to see a different.
Using a digital imager, some claim to have x2000 to x1000 fold linear range
especially if you are using fluorescent probes.
ELSA has high false rates that's why alpha-scan is developed. |
|
A******y
发帖数: 2041 | 3 I thought the ability for western blot to tell the difference has to do with
the linearity of the signal and maximum signal to noise ratio. If you use
film, you are limited by the film's linearity range which I "heard" is
between 2-10, so you need at least two fold differences to see a different.
Using a digital imager, some claim to have x2000 to x1000 fold linear range
especially if you are using fluorescent probes.
ELSA has high false rates that's why alpha-scan is developed. |
|
l**********1
发帖数: 5204 | 4 Appendix B
Mathematical Fundamentals
Differential Calculus
Linear Algebra
Probability
2012 Book final version link:
//basil2.math.uwaterloo.ca/~bingalls/AM382/Math_Modelling_in_Sysbio.pdf
Suggestions for Further Reading
A number of introductory calculus books are targeted towards applications in
the life sciences,
including Modeling the Dynamics of Life (Adler, 1998) and
Calculus for
Biology and Medicine
(Neuhauser, 2004).
A comprehensive introduction to partial differential
equations in provide... 阅读全帖 |
|
l**********1
发帖数: 5204 | 5 Appendix B
Mathematical Fundamentals
Differential Calculus
Linear Algebra
Probability
2012 Book final version link:
//basil2.math.uwaterloo.ca/~bingalls/AM382/Math_Modelling_in_Sysbio.pdf
Suggestions for Further Reading
A number of introductory calculus books are targeted towards applications in
the life sciences,
including Modeling the Dynamics of Life (Adler, 1998) and
Calculus for
Biology and Medicine
(Neuhauser, 2004).
A comprehensive introduction to partial differential
equations in provide... 阅读全帖 |
|
l**********1
发帖数: 5204 | 6 so what? if your found can solve solid wet or hard bio can’t solve non-
linear trend between
count level and variance by Deseq or other softs ?
pls refer,
01-16-2013, 12:40 AM #1
JesperGrud
Junior Member
Location: Odense
Join Date: Aug 2012
Posts: 5
DESeq and independent filtering
Hi everyone
I know this topic has been up a few times, but yet there is a question. So
the basic idea about filtering is that it is done unsupervised to remove
genes that are too lowly expressed to become signif... 阅读全帖 |
|
n**********g
发帖数: 196 | 7 1) I have no comments, but Dr. Chou has mentioned he was able to reproduce
it.
The biggest concern is not addressed by your answer.
2) DNA band and migration are easily alternated by sample ionic strength and
buffer.
No, I did not talk about band distortion. Read my question carefully. Fig 4a
and Fig 4e. No shift when cleavage sites are 30bp and then clear shift when
cleavage sites are closer.
3) Visualization/staining and capture will add non-linear effects to DNA
quantitation. Un-even chemical... 阅读全帖 |
|
发帖数: 1 | 8 I was a nucleic acid research scientist, but not working on gene editing any
more
. The reason I am involved is to present some basic knowledge on nucleic
acid
research.
{1) I have no comments, but Dr. Chou has mentioned he was able to reproduce
it.
The biggest concern is not addressed by your answer. }
Band shift sometimes can be seen due to sample ionic strength and buffer.
{ 2) DNA band and migration are easily alternated by sample ionic strength
and
buffer.
No, I did not talk about band dist... 阅读全帖 |
|
m*********D
发帖数: 1727 | 9 （CRISPR的背景资料请参阅Feng Zhang的“Genome engineering using the CRISPR-
Cas9 system”，Nature Protocols 8, 2281–2308 (2013)）
背景
我们是作一个与激素相关的肿瘤的。这个肿瘤的治疗方法已经相当完善，就是让身体里
不能生产激素或用一个小分子化合物抢占激素受体上激素的结合部位。但多年治疗后，
会有相当比例的病人产生抗药性。这个抗药性机理相当复杂。我们实验室一直致力于筛
选和现有药物不同的机理的小分子化合物，包括受体和DNA的结合或其他与受体相关的
pathway的inhibitors。前几年成功地筛选出了几个不同机理的化合物，个别在细胞水
平达到了nM的级别，动物测试也在10 mg/kg的水平。
最近在这个领域对抗药性肿瘤的研究发现，部分病人的抗药性是因为受体激素结合部位
发生了突变。一个或几个氨基酸的突变就能导致肿瘤细胞的生长不能被现有药物抑制。
所以，我们很想知道我们筛选出来的化合物对这些突变受体是不是也起作用。理论上来
说，我们的化合物不和激素竞争，突变不会影响我们化合物的效果。但只... 阅读全帖 |
|
c********n
发帖数: 225 | 10 对于韩春雨NBT文章无法重复的问题
需要从Figure 1 开始一步步看
作者的contributions
1. C.H. conceived the study and designed the experiments.
文章 “Story” “Idea” 的设计与发起人
2. X.Z.S. provided intellectual advice on the project and experimental
design.
实验设计， “One guide-faithful” -Figure 2d？
3. C.H. performed mammalian genome editing.
（Figure 4,5）
文章里让科研界最感兴趣的结果，也是无法被重复验证的主要问题
4. F.G. performed the BLAST search and the in vitro cleavage experiments.
（Figure 1, supplementary figure 2）
大肠杆菌E coli 表达的蛋白
NgAgo 基因编辑功能的初步 in vitro 数据支持
注... 阅读全帖 |
|
c********n
发帖数: 225 | 11 【再贴一次】
对于韩春雨NBT文章无法重复的问题
需要从Figure 1 开始一步步看
作者的contributions
1. C.H. conceived the study and designed the experiments.
文章 “Story” “Idea” 的设计与发起人
2. X.Z.S. provided intellectual advice on the project and experimental
design.
实验设计， “One guide-faithful” -Figure 2d？
3. C.H. performed mammalian genome editing.
（Figure 4,5）
文章里让科研界最感兴趣的结果，也是无法被重复验证的主要问题
4. F.G. performed the BLAST search and the in vitro cleavage experiments.
（Figure 1, Figure 2？，Figure 3a，supplementary figure 2）
大肠杆菌E coli 表达的蛋白
NgAg... 阅读全帖 |
|
x**u
发帖数: 21 | 12 Yes. I have tried decrease the minimum step time to 1e-9 and standard works.
However, the total increments will be 0.5/1e-9=500 million. I have to wait for
several days for the results to come out! And I have 125 similar cases to run.
Now I tried linear perturbation and it can give me results in ten minutes. It
makes me think that my part structure is linear or in my case, linear
perturbation can approximate the result.
thanks. |
|
h***o
发帖数: 539 | 13 move all the linear terms to the left hand side, and all the nonlinear term
to the right hand side.
use some initial guess to evaluate the right hand side. right hand side
becomes a constant.
then the equation becomes linear PDE...use whatever the way you like to
solve this linear PDE for one step, then update the
right hand side with the most recent result. repeat this process until
the solution converges.
I used this way to solve one elliptic pde with nonlinear terms last year. |
|
a*******o
发帖数: 129 | 14 【 以下文字转载自 CS 讨论区,原文如下 】
发信人: ayiyahiho (sddddddddddddddddd), 信区: CS
标 题: lineariation problem....help, please
发信站: Unknown Space - 未名空间 (Mon Jun 6 17:52:27 2005) WWW-POST
Hi, all,
I have a nonlinear constraint which i don't know how to linearize it.
it is in the form
x*y=z
x and y both have a bound.
Then, how can i linearize it? I hope I can achieve a LP or MILP form by
linearization.
Thank you very much. |
|
m*****1
发帖数: 5 | 15 【 以下文字转载自 Physics 讨论区 】
发信人: mph2011 (mph), 信区: Physics
标 题: Postdoctoral position available in medical physics
关键字: postdoc medical physics
发信站: BBS 未名空间站 (Mon Apr 4 09:41:30 2011, 美东)
A postdoctoral position in medical physics is available in the
Department of Radiation Oncology at University of Texas Southwestern
Medical Center at Dallas. The successful candidate will work on projects
related to image-guided radiation therapy (IGRT) using cone beam CT
(CBCT), PET and fluoroscopic imaging. T... 阅读全帖 |
|
f*****0
发帖数: 489 | 16 if you step out of the problem, you will notice that it is simply a
dataconversion issue: for any input (in your case, a 3-bit input), you will
need to find a corresponding 8-bit output.
your problem is very simple: it is absolutely linear and a 138 is a perfect
solution for that.
what if your input isn't perfectly linear? for example, you may do a current
to temperature conversion where the current isn't exactly linear to the
temperature?
In that case, you can use a very simple but highly effec |
|
i****e
发帖数: 170 | 17 pat pat
well, yes. the linearity usually can be find on the transconductance,
that's why we try to enlarge the constant range of the transconductance.
in a cascode structure, the transconductance is just the one on the
bottom transistor,which is very bad considering the linearity. usually
in order to achieve a high linearity, we use source degeneration,
feedback, current feedback, and so on, which decease the gain rapidly,
and might consume a lot power.
appears on
nonlinearity?
head |
|
j******g
发帖数: 1098 | 18 不是你之前问
Linear之类的公司,power部门毛利润60%左右.销售额几亿美元.
说说IP licensing这个盈利模式利润一般有多少? 总收入多少?
那么我就去查查Linear的数据啊,发觉Linear没有相关营收，那就找了一个在IP
licensing有代表性的公司dolby
您老不是喜欢跟人争数字正确性嘛，那我就给你找一个external会计师audit过的数据
，免得您老说我忽悠。
be professional啊！ |
|
g****t
发帖数: 31659 | 19 我听说是一家德国startup,真假不知道。
另外Linear在中国投入
不足，市场一直上不去。我觉得这条不改变，迟早死路一条。
你技术再好也没用。
据说Linear是最难进的。
另外Linear刚丢了Ipad的ＰＭＩＣ ｓｏｃｋｅｔ， 不知被那家抢走了。
大，这些公司也不是谁想进就能进的。 |
|
m*****1
发帖数: 5 | 20 【 以下文字转载自 Physics 讨论区 】
发信人: mph2011 (mph), 信区: Physics
标 题: Postdoctoral position available in medical physics
关键字: postdoc medical physics
发信站: BBS 未名空间站 (Mon Apr 4 09:41:30 2011, 美东)
A postdoctoral position in medical physics is available in the
Department of Radiation Oncology at University of Texas Southwestern
Medical Center at Dallas. The successful candidate will work on projects
related to image-guided radiation therapy (IGRT) using cone beam CT
(CBCT), PET and fluoroscopic imaging. T... 阅读全帖 |
|
d**g
发帖数: 1031 | 21 Linear Tech sue TI one patent, finally Linear Tech pay TI tens of Million
dollors.
Because Linear Tech violated TI more patents.
To go the the IP court, min cost is Million dollors even you win the case.
So that is why lawyer can make money.
That is why US company is very careful for IP issue. |
|
d********g
发帖数: 11948 | 22
oh不好意思我表达有问题
如果pass device的size需要很小而工作在linear区间，怎么去size这个pass device？
load current的spec是平均电流大小，而根据驱动电路，peak current可能达到很高的
值，这种情况下怎么去决定pass device的大小（size）？是根据平均最大电流来算
pass device大小还是根据peak current的值？pass device工作在linear区间有什么不
好的地方？ex，gain drop很多，compensation受影响（miller）。还有什么不好的地
方？对瞬时负载电流变化有没有什么影响？如果我output voltage要求dip在100mV范围
以内，pass device在linear region会有什么影响吗？ |
|
f******8
发帖数: 150 | 23 Under the supervision of VP, Chief Technical Officer, create and validate
large-signal non-linear models for range of RF power transistors. Initially,
models will be GaN but may eventually include LDMOS. The candidate must be
self-motivated and able to work independently with minimal supervision. The
candidate must be able to demonstrate a proven track-record in the
successful development of non-linear models for RF power transistors (
ideally GaN) with good agreement between modeled and measur... 阅读全帖 |
|
m*****1
发帖数: 5 | 24 【 以下文字转载自 Physics 讨论区 】
发信人: mph2011 (mph), 信区: Physics
标 题: Postdoctoral position available in medical physics
关键字: postdoc medical physics
发信站: BBS 未名空间站 (Mon Apr 4 09:41:30 2011, 美东)
A postdoctoral position in medical physics is available in the
Department of Radiation Oncology at University of Texas Southwestern
Medical Center at Dallas. The successful candidate will work on projects
related to image-guided radiation therapy (IGRT) using cone beam CT
(CBCT), PET and fluoroscopic imaging. T... 阅读全帖 |
|
j******d
发帖数: 70 | 25 偶不是学数学的,在看 a course in error correcting codes, p6 的Varchamov-Gilbert
bound 的证明实在是不理解.
there exists a binary linear code of length $n$, with at most $m$ linearly
independent parity checks and minimum distance at least $d$, if
\sum_{k=0}^{d-2}\binom{n-1}{k}<2^m
证明:建立一个 m*n 矩阵,其中任意 n-1 列 independent.第一列可以是任意非0的m位二
进制向量.now suppose we have chosen $i$ columns so that no $d-1$ are linearly
dependent. there are
c(i,1)+c(i,2)+...+c(i,d-2) lieaner combinations of these columns taken $d-2$
or fewer at a t |
|
g****e
发帖数: 1829 | 26 A lot of people have posted their answers. I think there is a point that one
should always bear in mind, (my personal opinion):
If X & Y are not independent and both are Normal distribution, we can draw a
conclusion: it must be LINEAR dependence, which means you can represent Y
~~~~~~
per follow up posts, this could be wrong (and I didn't really prove it).
in a linear form of X. the question then reduces to:
If X & Y are linearly dependent, is their sum still a normal? An |
|
n******t
发帖数: 4406 | 27 来自主题: Mathematics版 - 求答案书 数学上面linear algebra基本上是finite dimensional vector space，
而numerical linear algebra差不多是matrix computation.
但是很多工科的linear algebra课是这两种的混合物。 |
|
|
m*****1
发帖数: 5 | 29 【 以下文字转载自 Physics 讨论区 】
发信人: mph2011 (mph), 信区: Physics
标 题: Postdoctoral position available in medical physics
关键字: postdoc medical physics
发信站: BBS 未名空间站 (Mon Apr 4 09:41:30 2011, 美东)
A postdoctoral position in medical physics is available in the
Department of Radiation Oncology at University of Texas Southwestern
Medical Center at Dallas. The successful candidate will work on projects
related to image-guided radiation therapy (IGRT) using cone beam CT
(CBCT), PET and fluoroscopic imaging. T... 阅读全帖 |
|
R*j
发帖数: 248 | 30 I guess my question is: For those samples whose concentrations are beyond
the "Linear range" of standard curve in LC/MS/MS, can I dilute them to
linear range before analysis. Since linear curve makes more sense to most
people.
Apparently, another way is to establish a nonlinear standard curve which
covers the samples with high concentration.
Do you happen to know whether FDA agrees with nonlinear standard curve (e.g.
, quadratic equation)?
you |
|
G*******X
发帖数: 1028 | 31 validation 不是指用别的试验方法证实一下，而是指这个方法通过了一系列测试，证
明它测出的结果是可靠的。
凡是在pharmaceutical公司（包括CRO)做过analytical support的人，包括analytical
method development， QA,QC， 都知道这是什么。算是pharmaceutical analysis里
面的基本常识。
做validation，一般先要写一个 validation protocol，说明validation的步骤和
criteria。做完了要写valiation report。做起来不难，但一定要按部就班。
做液相色谱的，一般validation包括accuracy, precision, specificity, linearity,
robustness, detection limit,etc. 很多时候还做solution stability.
比如一个测API的concentration的LC method。做validation的时候，linearity一般测
的范围从50% 到150%。 假设100%... 阅读全帖 |
|
m*****1
发帖数: 5 | 32 A postdoctoral position in medical physics is available in the
Department of Radiation Oncology at University of Texas Southwestern
Medical Center at Dallas. The successful candidate will work on projects
related to image-guided radiation therapy (IGRT) using cone beam CT
(CBCT), PET and fluoroscopic imaging. The ideal candidate should have a
Ph.D. in Medical Physics, Biomedical Engineering, Electrical
Engineering, Physics, Mechanical Engineering or Computer Science.
Experience in medical imag... 阅读全帖 |
|
w*****1
发帖数: 220 | 33 Dear Readers,
We are happy to present the first five papers that are published in our
new, open access journal, Physical Review X. These papers, together with
others that are to be published in the coming weeks into September, will
constitute the first issue of the journal. Under PRX's open access
publishing model, they are free for you to read and use.
We are very encouraged by the breadth of their topical spread. It ranges
from the well-established field of atomic, molecular and optica... 阅读全帖 |
|
发帖数: 1 | 34 QUANTUM SPACE ELEMENTS (QSE)
INTRODUCTION
This is a non-mathematical proposal introducing a cosmological model that is
cyclic, deterministic and infinity-free. For the purpose of achieving this
objective a number of new concepts are introduced. The framework of this
proposal is structured over a number of assumptions, leading to various
considerations of cosmological concerns. Furthermore, verification methods
are suggested of both, observational and mathematical nature.
ASSUMPTIONS
Space is dis... 阅读全帖 |
|
c*****e
发帖数: 38 | 35 oh, for this, actually linear is the lower bound.
the proof is a reduction to search an element in an unordered array.
observe that one diagonal is unordered.
if you could do this in less than linear, you would be able to search an
element in unordered array in linear. |
|
c*****e
发帖数: 38 | 36 oh, for this, actually linear is the lower bound.
the proof is a reduction to search an element in an unordered array.
observe that one diagonal is unordered.
if you could do this in less than linear, you would be able to search an
element in unordered array in linear. |
|
t*******e
发帖数: 172 | 37 Suppose we can, take trace of this inequality.
you get
linear function <=linear function <=linear function
which can not be hold for all the x unless these three function are parallel
to each other.
However, if we make some restriction on the range of x, this problem is
trivial by simple le a_{0} negative enough and b_{0} big enough. |
|
j*****e
发帖数: 333 | 38 yes, vol is not supposed to be linear, but what they are talking abt is time
series modeling, not simply linear regression on some x variables, right?
linear |
|
H**r
发帖数: 10015 | 39 你学过线性代数吗。。。
这n个equations之间都是linearly independent的话，就是唯一解。
如果其中有两个是linearly dependent的，就只有n-1个linearly independent的
equations,这样会有一个未知数。。。 |
|
|
m******t
发帖数: 273 | 41 求纽约内部推荐 quant researcher, quant portfolio analyst
博士毕业 一年， 优化建模方向， 数值分析
optimization model design/analysis, linear/non-linear model building/solving
/analysis,
linear regression, statistics knowledge, robust portfolio optimization,
risk (VaR) modeling
commercial financial optimization engines (e.g. MSCI/Barra) principles and
applications
stress test, C++/R, basic SAS programming,
Any help would be appreciated !
Please email me : c******[email protected]
谢谢啦 ！！！ |
|
L*******t
发帖数: 2385 | 42 quant干的各种pricing，portfolio choice，帅哥你都没体会过其中的有趣啊。
我更喜欢这些。
最近才觉得empirical finance好玩的，喜欢上各种regression，主要也是喜欢其中的
数学的优美
你去google一下infinite dimensional linear regression，sublinear expectation
linear regression, continuous time linear regression看看。
JOF是金融方面文章，他们用的还是大数学大金工给他们提供的方法和路子啊 |
|
w**********y
发帖数: 1691 | 43 不要浪费时间去用过度复杂的model，model永远不是做出策略的最关键部分。
如果你这个idea能用,甚至都不需要用Kalman Filter,更用不着partical filter这些了
.
你要去想想kalman filter比linear regression增加的是什么feature. 而这些feature
实际应用的时候,你都可以用linear regression和选定local window和lookback
period去近似.
所以,唯一的建议,从linear regression开始做. 看看你做的过程中有哪些错误和观察
你上面说的transformation,没大看懂.如果你是 20*y = b*20*(x-1) + e
那你去交易你的spread当然能挣钱了. 你这样算出的spread已经被你人为加上drift了.
not tradable. |
|
L*******t
发帖数: 2385 | 44 本质上是MLE的estimation。或者换成Bayesian都行。
noise可以是很大一类分布。
然而效果depend on data和很多因素。
很简单的线性代数告诉我们如果你选择的factors构成了你的Y variable取值空间的一
组基
的话，那么线性关系是成立的。其他条件和情况你自己可以分析。。
如果从Taylor expansion的角度上来说，你是在把 Y = F(X)做泰勒展开然后取线性项
。效果是取决于函数F的性质的。如果F highly nonlinear，那么线性逼近没什么意义
了。
所以nonlinearity全部在noisy里面。
最新的技术是可以考虑model uncertainty的linear regression。作者的numerics表明
如果确实model有regime的话，用sublinear expectation linear regression效果会好
非常多。
所以用Linear regression要慎重啊。 |
|
c********s
发帖数: 123 | 45 When I was reviewing quantum physics, I cannot keep myself from noticing that
Quantum theory is such a linear theory. Is the microscopic world really
that linear. Is high energy phsics that linear?
hehe... |
|
d********g
发帖数: 294 | 46 For linear regression analysis as follows:
Y=X1 + X2 + error
Which of the following sentence is correct?
(1) Linear regression analysis for Y, X1, and X2.
(2) Linear regression analysis for X1, X2, and Y.
Should I write dependent variable at first (as written in (1)), or should I
write dependent variable at last (as written in (2))? I want to express it
in one sentence. Should I not use (1) nor (2), and use some other sentence
instead? If so, which sentence should I use?
Thanks! |
|
d******e
发帖数: 7844 | 47 严格的说，non linear dimensionality reduction也有一些，只不过本质上还是
linear的或者用linear来近似，直接按nonlinear研究的不多 |
|
R********n
发帖数: 519 | 48 对啊，呵呵，那些人脸上面的各种linear DR 或者近似linear DR，还有什么linear
subspace的方法，实在有点太多了，大同小异 |
|
s*r
发帖数: 2757 | 49 the p value will only tell you whether the correlation is significantly
different from 0. but the real question may be whether the correlation is
very close to 1
many ways can give you a pvalue. eg the pvalue from the regression
coefficient in a linear regression, and each way will have its problem, eg,
the normality assumption for linear regression if linear regression is used
in a small sample
.
other. |
|
q**********0
发帖数: 335 | 50 Thank you very much for your reply. But I still don't quite understand yet.
I used linear regression from Excel and got a equation and its R-square (R2=
0.239), Do you think this R2 value is a P value? My data set has 13 samples,
is it enough to do linear regression analysis? Except for linear regression
analysis, what else statistical method can be used (especially in Excel)?
,
used |
|
