search

BiomedicalMachineLearning / stLearn

A novel machine learning pipeline to analyse spatial transcriptomics data
Other
190 stars 25 forks source link

Run stSME clustering without HE #295

Closed rocketeer1998 closed 3 months ago

rocketeer1998 commented 3 months ago

Hi @quanaibn @BradBalderson @duypham2108 @jon-xu @xiao233333 , thanks for your awesome tool! I'd like to know whether I can run stSME without HE images at hand?

BradBalderson commented 3 months ago

Hi @rocketeer1998, stSME needs to have morphology data to do the normalisation, which comes from the H&E. So without it cannot do this. @xiao233333 would be the best person to as though.

xiao233333 commented 3 months ago

Hi @rocketeer1998, You are still able to use stSME normalisation. You can use spatial distance to normalise your data. Please see the attached code: 

st.spatial.SME.SME_normalize(data, use_data="raw", weights="physical_distance")
rocketeer1998 commented 3 months ago

Thanks for your clear explanation!