In general this data will be fetched as-needed by other loaders. However if you would like the transcripts to be back-filled to support auto-complete by users entering variants through the GraphKB client, it is sometimes useful to batch load these.
To complete a batch load you will first need to download the data
wget -O LRG_RefSeqGene.tab ftp://ftp.ncbi.nih.gov/refseq/H_sapiens/RefSeqGene/LRG_RefSeqGeneThis can then be loaded with the general file loader
node bin/load.js file refseq LRG_RefSeqGene.tab