Sometimes in a mass spectrometry experiment, it happens that
a few samples have been mislabelled. Mislabelling means that some replicates are
in the wrong sample group, and this results in some
samples having more (or less) replicates than the number intended by the experimentalist.
Apart from disparity in the number of replicates due to mislabelling, a few samples, e.g.
the quality control (QC) samples, are often assayed
several times. The aim is to analyze data with the same number of technical
replicates (in this case, duplicates) for every sample.
checkNo.replicates identifies samples with a disparate
number of replicates. The identified samples are treated as follows:
(i) The QC samples can be independently analysed to ascertain the reproducibility of the data.
(ii) The samples with no replicates are discarded from further analysis.
(iii) The samples with more replicates than expected, due to mislabelling (or otherwise), are
pre-processed using the function:
mostSimilarTwo which detects and discards
replicates which give conflicting peak information compared to the rest of the replicates.
Here, the two most similar replicates are treated as the correct replicates for
the sample in question.
checkNo.replicates(rawData, no.peaks, no.replicates)
Duplicate data in the same format as the raw data from the Biomarker wizard software.
The number of peaks detected by the Biomarker wizard
The number of replicates intended by the biologist.
It returns a vector whose elements are labels for samples with a disparate number peaks.
1 2 3 4 5 6 7 8 9