Create a dmatch class object. It is common to correct batch effects for more than two samples. We can run fastSVD with all those samples (as a list) at once. From this step on, we correct batch effects for pairwise samples at a time. We recommend to use the same dataset (sample) which has the most cells/the best quality as the reference sample, and use it to align the rest samples one at a time. The reference sample will be the latter in the pairwiseSamples.list.
CreatedmatchObject( pairwiseSamples.list, batch.id, project = "dmatchProject", PCA = PCA )
pairwiseSamples.list | Same as the input pairwiseSamples.list, or subset of the pairwiseSamples.list, for fastSVD. A list of pairwise samples for batch effects correction, each of which is a raw read count dataset. Alternatively, those samples can be the preprocessed and Lognormalized ones; then need to specify no preprocessing and no LogNormalize in the next step projection_to_reference_panel. |
---|---|
batch.id | Batch.ids which are used to denote those two samples in the previous fastSVD step. |
project | Project name (string). |
PCA | The output from fastSVD, i.e., the list returned by fastSVD function. |
Initializes the dmatch object. Return a dmatch class object which have slots storing pairwiseSamples.list, batch.id, PCA, and etc.