High % of reads with multiple alignments

[subxero]

Recently mapped some RNAseq reads with Tophat that were prepped from human whole blood, globin cleared, then prepped with Illumina Total stranded RNA with ribo zero gold (rRNA reduction and globin reduction + )

the fastqc run earlier shows high levels of duplications in these same libraries, >80% mapped but then around 70% of the mapped reads had multiple alignments.

Is this RNAseq data from these samples going to be usable? or pretty much garbage for any kind of RNA analysis?

multiple alignments possibly due to large number of rRNA and globin transcripts being sequenced? If not what are possible issues with the libraries?

[chenyudr]

We also met this multiple alignment situation. rRNA takes about <20% total reads and most multiple alignment sites are in one chromosome.