Tweet
Hi,
Lately we have been working with bacterial paired-end Illumina HISeq2000 RNAseq data. All the alignments were done in the core facility where we had our RNAseq performed. All I know is that the samples were sequenced in strand-specific manner, rRNA was removed, the length of the reads was 90 bp, and they were aligned using TopHat.
While analyzing the results I came across some problems. Depending on the way I try to visualize the alignments I keep on getting different results concerning the strand specificity. Therefore I have two main questions:
Question 1:
Attached are two files that show the same region viewed with IGV and Artemis. The difference is that Artemis was used to view the original BAM file (Artemis_XS_tagging.png), while IGV was used for viewing processed BAM files, which were first split according to the strand specificity (IGV after spliting.png). BAM files were split using the split.sh script. I’d like to know if the script is a reliable and which results to trust more, the ones seen with Artemis or the ones seen with IGV?
Question 2:
Viewing the BAM file with Artemis shows the antisense transcript originating always from the intergenic regions. It appears basically between each end every gene (Artemis_intergenic_regions.png). I assume that this is just an artifact due to technical problems. But where does this problem come from. How can I distinguish the true asRNA from these artifacts?
* IGV after spliting.png - Snapshot taken from IGV. The BAM file was first split using the script (split.sh) into two separate BAM files based on the strand specificity. Subsequently, both the files were converted into a TDF files. Two replicas of samples from bacteria grown at 22°C (blue bars) and two replicas of samples from bacteria grown at 37°C (red bars) were overlaid.
* Artemis_XS_tagging.png - Snapshot taken from Artemis. The original BAM file (bacteria grown at 37°C) was viewed with Artemis v16. Reads were colored by the RNAseq strand specific tag (XS). Alignment was viewed as a strand stack.
* Artemis_intergenic_regions.png - Snapshot taken from Artemis taken in the same way as Artemis_XS_tagging.png.
Lately we have been working with bacterial paired-end Illumina HISeq2000 RNAseq data. All the alignments were done in the core facility where we had our RNAseq performed. All I know is that the samples were sequenced in strand-specific manner, rRNA was removed, the length of the reads was 90 bp, and they were aligned using TopHat.
While analyzing the results I came across some problems. Depending on the way I try to visualize the alignments I keep on getting different results concerning the strand specificity. Therefore I have two main questions:
Question 1:
Attached are two files that show the same region viewed with IGV and Artemis. The difference is that Artemis was used to view the original BAM file (Artemis_XS_tagging.png), while IGV was used for viewing processed BAM files, which were first split according to the strand specificity (IGV after spliting.png). BAM files were split using the split.sh script. I’d like to know if the script is a reliable and which results to trust more, the ones seen with Artemis or the ones seen with IGV?
Question 2:
Viewing the BAM file with Artemis shows the antisense transcript originating always from the intergenic regions. It appears basically between each end every gene (Artemis_intergenic_regions.png). I assume that this is just an artifact due to technical problems. But where does this problem come from. How can I distinguish the true asRNA from these artifacts?
* IGV after spliting.png - Snapshot taken from IGV. The BAM file was first split using the script (split.sh) into two separate BAM files based on the strand specificity. Subsequently, both the files were converted into a TDF files. Two replicas of samples from bacteria grown at 22°C (blue bars) and two replicas of samples from bacteria grown at 37°C (red bars) were overlaid.
* Artemis_XS_tagging.png - Snapshot taken from Artemis. The original BAM file (bacteria grown at 37°C) was viewed with Artemis v16. Reads were colored by the RNAseq strand specific tag (XS). Alignment was viewed as a strand stack.
* Artemis_intergenic_regions.png - Snapshot taken from Artemis taken in the same way as Artemis_XS_tagging.png.
Comment