TargetSeeker-MS

          TargetSeeker-MS is a Bayesian Inference approach for drug target discovery, using protein fractionation coupled to mass spectrometry.
          It identifies proteins that are statistically significantly enriched via generating a noise-model using Bayesian Inference, which are measured with high reproducibility across a set of replicates.

          Sample input and output files are accessible here.

          Step 1. Number of control replicates (at least 4):


          Step 2. Number of drug replicates (recommended at least 2):


          Step 3. Number of sample fractions per replicate (at least 2):


          Step 4. Input non-normalized protein quantification values (e.g. spectral count, intensities) in the following tab-separated format (.txt file):
          The input file header must start with Protein and follow the following fraction format: xRyFz
              Where x is "C" for control replicate or "D" for drug replicate.
              Where y is the replicate number.
              Where z is the fraction number.
          The order of the columns after "Protein" does not need to be in order.
          However, each condition must have all replicates (1 to number of replicates) and all fractions (1 to number of fractions).
          Example:
          Protein    CR1F1    CR1F2    CR1F2    ...
          TO8B1.3        0      0        0
          W10C8.5       9     18      36
          F45D11.15    0      2        34

          See the sample input file above for a more detailed example.



          Step 5. k-nearest neighbor smoothing (integer greater than 0):


          Step 6. False-Discovery Rate threshold (number between 0 and 1):


          Step 7. Fold-change threshold (non-negative number):

This website is currently under development. If you encounter any problems, please contact mathieu.lavallee "at" uottawa "dot" ca

Powered by Yates Bioinformatics Team
Copyright @2016 by Yates Lab, Department of Chemical Physiology, The Scripps Research Institute