MAESTIMATE procedure

Estimates treatment effects from a two-colour microarray design (D.B. Baird).


Options

PRINT = strings
What to print (design, summary, monitoring); default desi, summ, moni

DYEBIASMETHOD = string
Whether to estimate dye bias effects (estimate, omit); default esti

SPREADSHEET = strings
What results to put in spreadsheets (estimates, df, rsd, dyebias, seestimates, tvalues, probabilities, contrasts, secontrasts, tcontrasts, prcontrasts); default esti, df, rsd, dyeb, sees, tval, prob, cont, seco, tcon, prco


Parameters

LOGRATIOS = variates or pointers
Log-ratios

PROBES = factors
Probes for the log-ratios

SLIDES = factors
Slides for the log-ratios

REDTREATMENTS = factors
Targets on red dye for slides

GREENTREATMENTS = factors
Targets on green dye for slides

CHECK = texts or variates
Slide ID's of the red and green treatments for a check matching the slide order with the labels or levels of SLIDE

XCONTRASTS = matrices
Contrasts to estimate

IDPROBES = texts
Saves the probe names for each output row

DF = variates
Saves degrees of freedom for t-values

RSD = variates
Saves the residual standard deviation

DYEBIAS = variates
Saves estimated dye swap bias effects

ESTIMATES = pointers
Saves the estimates

SEESTIMATES = pointers
Saves the standard errors of the estimates

TVALUES = pointers
Saves t-values of the estimates

PROBABILITIES = pointers
Saves probabilities for the t-values

CONTRASTS = pointers
Saves estimates of the contrasts

SECONTRASTS = pointers
Saves the standard errors of the contrasts

TCONTRASTS = pointers
Saves t-values for the contrasts

PRCONTRASTS = pointers
Saves probabilities for the contrasts


Description

MAESTIMATE estimate effects from the within-slide differences between targets (or treatments). This information is contained in the log-ratios. Usually, these log-ratios will have normalized using the MNORMALIZE procedure. MAESTIMATE uses analysis of variance with a pooled error across the targets (i.e. treatments) for each probe (or gene). The normalization of each slide effectively removes the block effects, so the log-ratios now reflect the differences between treatments on each slide, and the constant represents the dye-swap effect for the probe. The DYEBIASMETHOD option controls whether or not the dye biases are estimated, and the XCONTRASTS parameter allows you to specify a matrix defining contrasts to estimate between treatments.

   The log-ratios are supplied by the DATA parameter. If these are in a single variate, the SLIDE parameter should supply a factor to index the slides, and the PROBES parameter should index the probes or genes. Alternatively you can supply a pointer containing a variate for each slide. The slides factor is then not required; if it is given it should just have one entry for each slide in the order of the variates in the pointer. The PROBES factor is that for a single slide, and all slides must have a common layout.

   The REDTREATMENTS parameter should supply a factor to indicate the target assigned to the red dye. This is assumed to be the channel on the top of the log-ratios. This factor must have the same number of values as the number of levels of the Slides factor. Similarly, the GREENTREATMENTS parameter should supply a factor to indicate the target assigned to the green dye. The CHECK parameter can supply a text or variate identifying the slide in each unit of the REDTREATMENTS and GREENTREATMENTS factors. This can then be used to check that these units match the slides according to the labels or levels of the SLIDE factor. If the labels of the slides and check factor match, but are in a different order, the treatment factors will be sorted into the correct order, and a warning is given.

   The other parameters allow you to save results from the analysis, and the SPREADSHEET option allows these to be put into GenStat spreadsheets.

 

Options: PRINT, DYEBIASMETHOD, SPREADSHEET.

Parameters: LOGRATIOS, PROBES, SLIDES, REDTREATMENTS, GREENTREATMENTS, CHECK, XCONTRASTS, IDPROBES, DF, RSD, DYEBIAS, ESTIMATES, SEESTIMATES, TVALUES, PROBABILITIES, CONTRASTS, SECONTRASTS, TCONTRASTS, PRCONTRASTS.