What 's difference between genome strand and transcription direction?

[sixguns]

My RNA-seq data is non-strand specific.
After mapping data to genome by bowTie,
I confused meaning in results file : strand '+' or '-' ?

for example:
readscore1 + chr1 10000 tag1 ...
readscore2 - chr1 10000 tag2 ...

Does it mean that the positions as above are in the same place of genome
and they are complementation?

Is it related to transcription direction?

[steven]

Hi,
If your protocol is not strand-specific (oriented), then you cannot conclude anything from the "strand" field, this information is not reliable. The aligner takes a sequence in input, and reports the strand it has been aligned to, but each sequence can come from the plus strand as well as the minus one, there is no way to be sure.
Unless the protocol is strand-specific, RNA molecules from the same transcription unit will typically produce reads in both directions although they are transcribed in the same direction.
Guys, please correct me if i'm wrong.

[sixguns]

Thank you Steven!
Here is the example data (non-strand specific):
tag1 - chr1 1747195 TGGAGTCTGTGCCCCAGTGCATGGCGTAGATCTTGGCCAGGTGCCCCCGCAGTGTCCTCCTCGTGCGCATTTGGA score1 0
tag2 + chr1 1747226 CTTGGCCAGGTGCCCCCGCAGTGTCCTCCTCGTGCGCATTTGGATTCTTCCCACTGGGTCGATGTTGTTTGTGAT score2 0

You can see that the underline sequences of tag1 and tag2 is the same. I don't know why result of biwtie show strand of tag1 is '-', it seems that '+' is right.

[steven]

Looks like the strand information is definitely not reliable, as i thought. I have not used bowtie myself, but you could post the command line you used and experts may see the reason. Did you use the "-nrc" option for instance (see this thread)?
Any bowtie user around?