I have recently run a relatively obscure genome through RepeatMasker and RepeatModeler, and produced a bedfile whereby I have all the instances of a TE annotated on all scaffolds.

This is what is looks like:

scaffold1       4       209     LINE/Penelope   0       +
scaffold1       429     470     Unknown 0       +
scaffold1       1027    1083    Low_complexity  0       +
scaffold1       1444    1484    Low_complexity  0       +
scaffold1       1713    1803    Low_complexity  0       +
scaffold1       1900    1920    Simple_repeat   0       +
scaffold1       3464    3538    Simple_repeat   0       +
scaffold1       4994    5045    Simple_repeat   0       +
scaffold1       7516    7541    Simple_repeat   0       +
scaffold1       10543   10827   DNA/hAT-Charlie 0       -
scaffold1       10823   11007   DNA/hAT-Charlie 0       -
scaffold1       11430   11506   Low_complexity  0       +
scaffold1       11562   11701   LINE/Penelope   0       -
scaffold1       12150   12181   Low_complexity  0       +

I also have a transcriptome, which has the TSS, TES, and exon coordinates annotated. It looks like this:

SCAFFOLD        TSS    TSS 
scaffold1       1206    7550    transcript_2.1  .       +        transcript    .       gene_id "transcript2.1"; transcript_id "transcript2.1";
scaffold1       1206    1257    transcript_2.1  .       +        exon    .       gene_id "transcript2.1"; transcript_id "transcript2.1";
scaffold1        6257    7550   transcript_2.1  .       +        exon    .       gene_id "transcript2.1"; transcript_id "transcript2.1";

If I wanted to annotate the transcriptome for these particular TEs, would it be okay to do a simple bedtools intersect between the annotated genome file and the transcriptome file?

Can there be a biological situation where you have a transcript which is a protein-coding gene but within one of its exons also has an annotation as an LTR for example?