CellMix 1.6

Complete Deconvolution using Digital Sorting Algorithm (DSA)

Description

The method “DSA” implements the Digital Sorting Algorithm (DSA) proposed by Zhong et al. (2013), which performs complete gene expression deconvolution using a set of marker genes only.

Usage

gedAlgorithm.DSA(..., data = NULL, exact = TRUE, log = NULL)

Arguments

...
extra parameters DSAproportionspassed to mlsei.
data
markers as MarkerList object.
exact
logical that specifies if one should impose a sum-to-one (TRUE) or sum-to-less-than-one (FALSE) constraint on the proportions.
log
indicates if the data are in log-scale or should be converted to linear-scale. This is relevant because the DSA algorithm assumes that the input mixed data are in linear scale (i.e. not log-trasnformed). If NULL, then data's scale is detected by link{is_logscale} and conversion to linear-scale is performed if necessary. If TRUE the data is exponentialised (using log base 2). If FALSE the data is left unchanged (the detected log scale is displayed in verbose mode). If a number, then it is used as the log base to exponentialise the data.

References

Zhong Y, Wan Y, Pang K, Chow LM and Liu Z (2013). "Digital sorting of complex tissues for cell type-specific gene expression profiles." _BMC Bioinformatics_, *14*(1), pp. 89. ISSN 1471-2105, , .

Examples


## Example on dummy/random data
x <- rmix(3, 100, 20)
# markers
ml <- getMarkers(x)

# deconvolve using DSA (quadratic programming)
res <- ged(x, ml, 'DSA', verbose=TRUE, log=FALSE)

##   Using ged algorithm: "DSA"

## Note: method with signature 'MatrixData#MarkerList#ANY' chosen for
## function 'nmf', target signature 'ExpressionMix#MarkerList#character'.
## "ExpressionSet#ANY#ANY" would also be valid

##      Estimating basis and mixture coefficients matrices from marker features [DSA]
##      Using 15/15 markers to estimate cell proportions: 
##  CL_1 CL_2 CL_3 
##     5    5    5 
##       Checking data scale ...       NOTE [log]
##       Converting data to linear scale ...       SKIP
##       Computing proportions using DSA method ...       OK
##       Estimating basis matrix from mixture coefficients [qprog]
##       Not using any marker constraints
##   Timing:
##    user  system elapsed 
##   1.432   0.056   1.489 
##   GED final wrap up ...   OK

# plot against true data
profplot(x, res)

plot of chunk unnamed-chunk-1 

profplot(t(basis(x)), t(basis(res)), legend=FALSE)

plot of chunk unnamed-chunk-1 


## Don't show: 
    stopifnot( nmf.equal(res, ged(x, ml, 'DSA', log=FALSE), identical=FALSE) )
## End Don't show