What types of experimental data values can be imported into Pathway Studio?
1.) What are the basic types of numeric data supported for import by Pathway Studio
Pathway Studio supports import of signal intensity data, log-intensity, ratio, and log-ratio data.
2.) What base does Pathway Studio use for logarithmic data?
All logarithmic data in Pathway Studio is base 2.
3.) Please define these numeric data types.
Signal data consists of positive integer values, typically ranging from a value near zero to a value of several thousand. Signal values should have no negative numbers.
Log signal is a log base 2 transformation of the signal data. Taking the logarithm of the signal data tends to compress values so that very large and very small values will not have an overwhelming effect on experimental results.
Ratio data consists of a comparison between two signals, or simply “Signal 2 divided by Signal 1.” Ratio data can arise from 2-color assays or from comparison of two sample groups. Ratio values are never negative and range from zero to infinity. Ratio data is frequently referred to as “Fold-Change.”
Log ratio data consists of the log of the fold-change or ratio data. Taking the log of the ratio data tends to compress values. Log ratio data consists of a continuum of values between -infinity and +infinity. A value of zero indicates no change between sample2 and sample1. Values less than zero indicate a downward modulation of sample 2 vs. sample 1. Values greater than zero indicates an up regulation in sample 2 vs. sample 1.
4.) Does Pathway Studio support import of signed fold-change?
No. Import of signed fold-change is not currently supported by Pathway Studio. You should transform this data to log-ratio prior to import in Pathway Studio. Signed fold-change is a very visually appealing format. Upregulated signals are indicated by numbers ranging from 1 to infinity. Down regulated signals are indicated by numbers ranging from -1 to –infinity. No change is indicated by a zero.
5.) How can I transform signed fold-change data for import into Pathway Studio
The following rules indicate how to change signed fold-change data into log-ratio data.
Signed Fold-Change Range | Equation |
1 to +infinity | Log_base_2(signed fold-change) |
0 | 1 |
-1 to -infinity | Log_base_2(-1/ signed fold-change ) |
You can use Excel to make this transformation as follows:
1.) Beginning with an Excel table of signed fold changes
A | B | C | |
1 | probe set | FC 30 min | log-ratio |
2 | 245047_at | 2.88 | |
3 | 245048_at | 3.03 | |
4 | 245049_at | 1.5 | |
5 | 245050_at | 2.59 | |
6 | 245051_at | -1.46 | |
7 | 245052_at | -1.47 | |
8 | 245053_at | -1.04 | |
9 | 245054_at | -1.19 | |
10 | 245055_at | -1.09 |
2.) Copy the following equation into the cell C2
=IF( B2 > 0, LOG(B2,2), (IF(B2<0, LOG(-1/B2,2),0)))
3.) Propagate the formula from C3 through C10.
A | B | C | |
1 | probe set | FC 30 min | log-ratio |
2 | 245047_at | 2.88 | 1.526069 |
3 | 245048_at | 3.03 | 1.599318 |
4 | 245049_at | 1.5 | 0.584963 |
5 | 245050_at | 2.59 | 1.372952 |
6 | 245051_at | -1.46 | -0.54597 |
7 | 245052_at | -1.47 | -0.55582 |
8 | 245053_at | -1.04 | -0.05658 |
9 | 245054_at | -1.19 | -0.25096 |
10 | 245055_at | -1.09 | -0.12433 |
6.) What display formats are available for experiment data?
The Experiment Data Viewer will display sample values in the same format for which they were imported in. For note, Affymetrix .CEL values will be displayed in log format.
Differential expression values can be displayed in fold-change, log-change, signed fold-change, and absolute fold-change.
Fold-change is the ratio of sample2 vs. sample1, or sample2 / sample1.
Log-change is the same as the log of the fold-change.
Signed fold-change applies the following transformation to the log-change:
Fold change value | Transformation |
FC >1.0 (sample2 > sample1) | Signed Fold-change = Fold-Change |
FC = 1 (sample2 = sample1) | Signed Fold-change = 0 |
FC < 1.0 (sample 2 < sample1) | Signed Fold-change = -1 / Fold-Change |
Absolute Fold-Change is the absolute value of the signed fold-change. In this way sample values, which are modulated equally up and down, are shown the same.
7.) What kinds of operations are allowed on experiment data
Sample data can be used to color network diagrams and to show on expression line and heat-map diagrams.
(see Figure 1 at the top of this entry)
For differential expression values you can run various algorithms, including GSEA, SNEA, Find Networks Enriched with Selected Genes. You can also run these algorithms on imported log-ratio and ratio (i.e. fold-change) data.
