Trend Analysis


Select a file Example
The table contains gene expression value, in which the first column is gene IDs and the others are expression information. The sample order in this table will determine the trend order in the result.
*
All gene will be clustered into profiles, and the number of profiles will be no more than this threshold. No more than 20 profiles is suggested because too many profiles will fragmentize the result and are hard for interpretation.
*
All time series will be transformed so that the time series starts at 0. This can be done in one of three ways based on the option selected to the left. Given a time series vector of values for a gene (v_0,v_1,...,v_n) the options are:
Option 1. 'Log normalize data' − the vector will be transformed to (0,log₂(v_1)−log₂(v_0),...,log₂(v_n)−log₂(v_0)). Note that any values which are 0 or negative will be treated as missing.
Option 2. 'Normalize data' − the vector will be transformed to (0,v_1−v_0,...,v_n−v_0)
Option 3. 'No normalization/add 0' − a 0 will be inserted transforming the vector to (0,v_0,v_1,...,v_n)
*
The value will determine profiles of genes enrichment significantly after permutation test. The default value is 0.05.
*
Greater than zero
After transformation (Log normalize data, Normalize data, or No Normalization/add 0), if the absolute value of the gene's largest change between any two time points is below this threshold, then the gene will be filtered. The default value is 2.
Select a file Example
*ProVIP free
*VIP free
*0 left in the current account
Trend analysis tools use tutorials and parameter explanation

Function:
Gene expression pattern analysis is used to cluster short time series expression data. The kernel program of this analysis is STEM tool (Short Time-series Expression Miner:http://www.sb.cs.cmu.edu/stem/)


Input:
Input tab delimited text files with headers. If your data is not TAB delimited, you can use Excel to convert it.
The table contains gene expression value, in which the first column is gene IDs and the others are expression information. The sample order in this table will determine the trend order in the result. Gene annotation file contains gene description, pathway annotation, GO annotation, etc, with gene IDs as the first column.


Parameters:

① Maximum number of Model Profiles:All gene will be clustered into profiles, and the numbers of profiles will be no more than this threshold. No more than 20 profiles is suggested because too many profiles will fragmentize the result and are hard for your interpretation.


②Data normalization:
  All time series will be transformed so that the time series starts at 0. This can be done in one of three ways based on the option selected to the left. Given a time series vector of values for a gene (v_0,v_1,...,v_n) the options are:
Option 1.Log normalize data'− the vector will be transformed to (0,log₂(v_1)−log₂(v_0),...,log₂(v_n)−log₂(v_0)). Note that any values which are 0 or negative will be treated as missing.
  Option 2.'Normalize data' − the vector will be transformed to (0,v_1−v_0,...,v_n−v_0)
  Option 3.'No normalization/add 0' − a 0 will be inserted transforming the vector to (0,v_0,v_1,...,v_n)


③P value threshold of significant profiles:The value will determine profiles of genes enrichment significantly after permutation test. The default value is 0.05.


④Minimum absolute expression change: After transformation (Log normalize data, Normalize data, or No Normalization/add 0), if the absolute value of the gene's largest change between any two time points is below this threshold, then the gene will be filtered. The default value is 2.


⑤Add annotation file: This is an alternative parameter. The first column is gene IDs and the others are gene annotation as your customized. The information provided here will be add to the output table and this will help you interpret the result.


Output:

①all_profile.xls:A table containing cluster results of all profiles.

②all.xls:the gene expression information of your input file.

③trend_all_by_gene_number.png:The profiles image ordered by gene number in each profile.

④trend_all_by_pvalue.png:The profiles image ordered by enrichment p value.

⑤Profile*.png:detailed cluster image of each single profile.

⑥Profile*.xls:a table containing culster results of each single profile.

Example file: example file of the table file    example file of description



Representative output result:

1. the figure of trend_all_by_pvalue.png

生物云平台


2.all.xls The table contains gene id and expression value, and gene description if you have provided in the “add description” section.



3. all_profile.xls The table contain cluster result of all profiles as follow.

生物云平台


4. profile[0-n].png,profile[0-n].xls detailed cluster image of single profile as follow.

生物云平台