Integrating Data QCed in Seurat


I am trying to load in a file from R that I have QCed using Seurat. I want to try to integrate through scanpy but I am concerned about what data is present in my Seurat object that may have an effect on the integration process. I have the following data present in the initial loom file when I load it in:

AnnData object with n_obs × n_vars = 36927 × 16872
obs: ‘ClusterID’, ‘ClusterName’, ‘SCT_snn_res_0_8’, ‘nCount_RNA’, ‘nCount_SCT’, ‘nFeature_RNA’, ‘nFeature_SCT’, ‘orig_ident’, ‘percent_mito’, ‘percent_mt’, ‘seurat_clusters’
var: ‘Selected’, ‘vst_mean’, ‘vst_variable’, ‘vst_variance’, ‘vst_variance_expected’, ‘vst_variance_standardized’
layers: ‘norm_data’, ‘scale_data’
If I remove the sct assay in seurat and look at that file I appear to see more genes I believe because the processing involved in generating the sct assay has removed genes present in less than 5 cells. This is the information I get from the object with no sct assay:
AnnData object with n_obs × n_vars = 36927 × 33694
obs: ‘ClusterID’, ‘ClusterName’, ‘SCT_snn_res_0_8’, ‘nCount_RNA’, ‘nCount_SCT’, ‘nFeature_RNA’, ‘nFeature_SCT’, ‘orig_ident’, ‘percent_mt’, ‘seurat_clusters’
var: ‘Selected’, ‘vst_mean’, ‘vst_variable’, ‘vst_variance’, ‘vst_variance_expected’, ‘vst_variance_standardized’
layers: ‘norm_data’
I wasn’t sure if any of this data being present would cause an issue with trying to integrate my data? I have run integration and do appear to have some batch effect present. I have tried to remove the scale_data and norm_data layers to rerun normalization using scanpy but it doesn’t seem to resolve my issues. I am a bit new to python and don’t fully understand how these data types may be used by scanpy in normalization and integration. Any help or advice anyone could think to offer would be very much appreciated.

Thank you