2 |
26 Feb 07 |
jari |
1 |
<html> |
2 |
26 Feb 07 |
jari |
2 |
<body bgcolor = "#FFFFCC"><basefont face = "Arial"> |
2 |
26 Feb 07 |
jari |
3 |
<h1>Iterative Linear Regression Normalization</h1><h2>Parameter Information</h2> |
2 |
26 Feb 07 |
jari |
4 |
<hr size = 10> |
2 |
26 Feb 07 |
jari |
5 |
|
2 |
26 Feb 07 |
jari |
6 |
<h2>Normalization Mode</h2> |
2 |
26 Feb 07 |
jari |
7 |
There are two possible normalization modes available. |
2 |
26 Feb 07 |
jari |
8 |
<h3>Block mode</h3> |
2 |
26 Feb 07 |
jari |
9 |
Block mode selects to traverse the |
2 |
26 Feb 07 |
jari |
10 |
data block by block and performs the normalization within each block. |
2 |
26 Feb 07 |
jari |
11 |
Using this mode data variation between blocks is ignored. |
2 |
26 Feb 07 |
jari |
12 |
<h3>Global Mode</h3> |
2 |
26 Feb 07 |
jari |
13 |
Using this mode all data enters into one global normalization. In this |
2 |
26 Feb 07 |
jari |
14 |
mode the normalization of all data points is influenced by all other data points |
2 |
26 Feb 07 |
jari |
15 |
regardless of block membership. |
2 |
26 Feb 07 |
jari |
16 |
<br> |
2 |
26 Feb 07 |
jari |
17 |
<h2>Standard Deviation Range</h2> |
2 |
26 Feb 07 |
jari |
18 |
On each iteration a linear regression is calculated for log(Cy5) vs. log(Cy3) based on |
2 |
26 Feb 07 |
jari |
19 |
a subset of the data. Initially this subset is the entire data set. |
2 |
26 Feb 07 |
jari |
20 |
On each iteration points which fall outside the number of standard deviations indicated |
2 |
26 Feb 07 |
jari |
21 |
from the regression line are omitted for the next iteration. On each subsequent |
2 |
26 Feb 07 |
jari |
22 |
iteration the R<sup>2</sup> is taken until it converges and for each iteration the data set |
2 |
26 Feb 07 |
jari |
23 |
for this regression is reduced by selecting only points which fall within the provided |
2 |
26 Feb 07 |
jari |
24 |
standard deviation range. |
2 |
26 Feb 07 |
jari |
25 |
</basefont> |
2 |
26 Feb 07 |
jari |
26 |
</body> |
2 |
26 Feb 07 |
jari |
27 |
</html> |