./filter_low_expr_transcripts.pl

Question

[kishorssf91]

Hi, I have to assmble and annotate a transcriptome only from one tissue. After asembly, Based on RSEM during filtering step i have used "filter_low_expr_transcripts.pl" script from trinity package. This scripts requires matrix with TPM or FPKM. Though i have a table (RSEM.isoforms.results) after running align_and_estimate_abundance.pl containg IDs, lengths, expected_counts, TPM, FPKM, IsoPct, when i used this result to generate a matrix in abundance_estimates_to_matrix.pl script, it probably gave me the result only based only on "expected_counts". But my questions are:

1. As for filtering it requires FPKM or TPM value and i have received only expected count values (possibly due to having only one sample), is it technically relevent to filter based on expected count values? And what would be the possible cut off value if i have to filter based on "expected_count"?
2. Will it be fine if i filter manually based on upstream FPKM value?

I am attaching pictures for your convenience. Your help and sugggestions are expected. Thanks