integrating scATAC-dataset across species using motif orthlogs

[ronghai-li]

mentioned in the literature：
Instead, we found that clustering based on TF topic enrichment could reasonably align brain subclasses between species.

However, I have in the code 06.integration/inte.coembed.motif.R

#------------------------------
# reference-based clustering
quy.se@meta.data$queryCluster <- quy.se@meta.data[, quyAttr]
ref.se@meta.data$refCluster <- ref.se@meta.data[, refAttr]

quy.se@meta.data$set <- "query"
ref.se@meta.data$set <- "ref"

# get PCA based on motif z-score
ref.feature <- rownames(ref.se[["RNA"]])
ref.se <- RunPCA(ref.se, features = ref.feature)
#ref.se <- FindNeighbors(ref.se, dims = 1:30)

quy.feature <- rownames(quy.se[["RNA"]])
quy.se <- RunPCA(quy.se, features = quy.feature)
#quy.se <- FindNeighbors(quy.se, dims = 1:30)

se.lst <- list(ref=ref.se, query=quy.se)

# find anchors
se.features <- intersect(rownames(ref.se[["RNA"]]), rownames(quy.se[["RNA"]]))
se.anchors <- FindIntegrationAnchors(object.list = se.lst, anchor.features = se.features)
se.combined <- IntegrateData(anchorset = se.anchors)

# Perform an integrated analysis
DefaultAssay(se.combined) <- "integrated"

From the code you provided and the description of the problem, there really is no obvious code segment that shows direct cross-species integration using motifs or motif orthologs.What did I miss?