lncRNA differential expression using RNA-seq

[LP_SEP23]

Hi All,
I have been doing RNAseq for differential expression using standard reference genome mapping and using EdgeR or DESeq to detect differential expression with no problem. I also have extensive microarray and qPCR data for the same samples, so a healthy dose of positive controls for known genes.

While validating one of my candidate differentially expressed genes (which was a microRNA) y realized that what was actually differentially expressed was novel lncRNA just upstream. After looking through the data, I have found several instances in which I have what appears to be lncRNAs which are greatly induced by my treatment.
My questions is: is there a way to scan in an unbiased way for yet to be annotated long non-coding genes that are differentially expressed between treatment and control? Would a de-novo assembly strategy like Trinity, followed by edgeR or DEseq work?

I am working with mouse tissue, BTW

any suggestions will be greatly appreciated

Lucia

[Dario1984]

The problem will always be proving that it is non-coding. Absence of proof is not proof of absence.