Fumiko Study - Data Preprocessing from Biostatistics Shared Resource Core

• Start Date: 14 June 2005
• Research Drive Directory: /Taguchi/050304/Binning
• Data Type: MassSpec
• Notes
  • uses both C# and SAS to calculate QC for testing purpose
  • folders dataExplore, Bining, and proTS are data sets have binned with old methods (with or without intensity)
  • folder newMethod are data sets that have new methods

• Status:

Output

1. Powerpoint report: /Taguchi/050304/Binning/deliverable report.ppt
2. cluster files: /Taguchi/050304/Binning/QC/cluster
3. Quality Control result: /Taguchi/050304/Binning/QC/
4. Shuo winner list(Unequal Variance T value) : /Taguchi/050304/Binning/WFCCM2 winner.xls
5. proTS winner list(Unequal Variance T value) : /Taguchi/050304/proTS/WFCCM2 winner.xls
6. dataExplore winner list(Unequal Variance T value) : /Taguchi/050304/dataExplore/WFCCM2 winner.xls

Change Log

Analysis Requests

1. QC serum control (by date, by subject, one subject cross all others)
2. Italy real data set(based on average) - WFCCM, 10 fold
3. WFCCM Japanese real data set
   1. pre-treatment alone (based on average) - Kanazawa, JFCR, Kanazawa+JFCR
   2. 10 fold on combined JFCR&Kanazawa
   3. pre&post : Kanazawa, JFCR.
      1. WFCCM -mass charge both pre&post have intensitity, and 20% of them have intensitity("and" condition- PR >= 20% && PD >= 20% && SD >= 20%), others are missing
      2. Fisher - only for YN or NY situation (pre - with intensitity, post - without intensitity. Or, pre - without intensitity, post - with intensitity), and 20% AND condition, others are missing
   4. training and testing
      1. training: Kanazawa testing: JFCR
      2. training: JFCR testing: Kanazawa

Contact Info:

• Primary Study Contact Person: Fumiko 6-3819

Analyst

• JoanZhang