Human wrapper script

WrapperAIBSHuman.m

@@ -0,0 +1,42 @@
+%This wrapper script runs through the figures and results from the MISS
+%manuscript, and regenerates them in order, beginning with the basic
+%pipeline results using the AIBS human scRANseq dataset
+
+%FILEPATH
+matdir = '/Users/justintorok/Documents/MATLAB/MISS/Human-related/RawDataFiles'; %define directory to draw from and save data to
+
+%LOADING INITIAL INPUT DATA
+load([matdir filesep 'AIBS_Human_scRNAseq_datastruct_subclass.mat'],'subclass_struct');
+load([matdir filesep 'abagen426_1mm_datastruct.mat'],'regexpr_struct');
+genevct = subclass_struct.averages.genevct;
+classkey = subclass_struct.cell_types;
+gene_names = subclass_struct.gene_names;
+regvgene = regexpr_struct.regvgene;
+regvgene_sums = sum(regvgene);
+regvgene_sums = repmat(regvgene_sums,size(regvgene,1),1);
+regvgene = regvgene ./ regvgene_sums;
+clear regexpr_struct subclass_struct
+
+%MRx3 GENE RANKING USING scRNAseq FROM AIBS
+lambda = 90; %percentile of genes to exclude from MRx3 ranking based on projection error added
+geneinds =  MRx3_Selector_Prefilter(genevct,regvgene,size(regvgene,2),lambda,0); %generating MRx3 gene indices
+save([matdir filesep 'MRx3_l90_human_geneinds.mat'],'geneinds');
+
+%DEFINING MAPPING PARAMETER INPUTS
+ng_param_list = [100:100:1000, size(regvgene,2)]; %values of nG to test and map, going through genes in MRx3 ranked order
+% ng_param_list = [50:10:1000, 1050:50:2000, 2100:100:3900, 4000:500:14500, size(regvgene,2)];
+missmethod = 'MRx3'; %gene ranking/subsetting method as a label, options are 'MRx3' and 'none'
+infmethod = 'inv+res'; %inversion method between E and C*D, options are 'inversion', 'inv+res' to also get residuals, and 'corr' for correlation mapping
+savename = 'CellDensity_Human_allnG.mat'; %set name of file to be saved
+
+%GENERATING MAPS, RESIDUALS, METADATA, & ELBOW INDEX/CURVE
+outstruct = Cell_Density_Outstruct_Human(genevct,regvgene,... 
+    gene_names,ng_param_list,lambda,missmethod,infmethod,...
+    [],matdir);  %getting outstruct of maps and residuals across nG            
+
+%DEFINING ELBOW
+makefig = 1; %binary flag whether or not to make elbow curve plot
+elbowind = elbow_selector_human(outstruct,makefig); %getting elbow index value and generating elbow curve
+
+save([matdir filesep savename],'outstruct','ng_param_list','lambda',... %saving cell mapping output
+        'missmethod','infmethod','geneinds','elbowind');