Nanostring-Biostats / CosMx-Cell-Profiles

Other
5 stars 1 forks source link

Not all genes from 1K CosMx in references? #2

Closed roanvanscheppingen closed 6 months ago

roanvanscheppingen commented 6 months ago

Given the mouse dataset and a user-generated Cosmx seurat I run Insitutype. My expression matrix has a length of 1000 (genes)

Your references have a length of 950 (genes).

The following genes in the count data are missing from fixed_profiles and will be omitted from clustering: Abl2,Acaca,Adcy5,Atg10,Bmpr1b,Cdc14b,Cplx1,Csmd3,Ddt,Dnah6,Dync1i1,Dynll1,Entpd2,Fnip2,Icmt,Lama2,Lgi2,Lrrc4c,Maml3,Masp1,Neo1,Neu4,Pde1c,Pde7b,Pgls,Phactr1,Pias1,Pik3c3,Pla2g4c,Plcl1,Pmch,Prps2,Ptprj,Rspo2,Sh3gl3,Slc24a3,Slc39a10,Snhg11,Sos2,Sstr3,Tab2,Tacr3,Thsd4,Tspan3,Ulk2,Unc5d,Uqcrq,Usp34,Vav3,Zfyve9

Why are these genes not in the profiles? And given that they are not exactly the same (something eluded to in the readme), should the reference profile then be updated for use in InSituType?