New reagents & software for GAIIx (an maybe IIe?)

[kmcarr]

Just heard this from our FAS a couple of minutes ago. They will very soon announce new reagent kits and software for the GAIIx. These will allow sequencing runs up to 2 x 151 cycles and slightly higher cluster densities on the IIx. (I don't know if this applies to the IIe also.) The software will be SCS 2.8. Oddly they are skipping 2.7. He did relate one rather surprising detail about this software; it will not allow saving of images at all! (other than thumbnails). This means no more possibility to do offline base calling. We routinely save images for 30 days, to be safe, and I just recently was able to salvage a run which had an instrument problem during cycle #8. We merely extended the run by 10 cycles, I then ran OLB on cycles 11-46.

I'm slightly unhappy about this. How do others feel about no longer having the possibility to perform offline base calling?

[Efortier]

Hi,

The SBS 2.8/RTA 1.8 update (for GAII, GAIIx, GAIIe) is already available on icom.illlumina.com. In the release notes it is written that you no longer can save images if you are runnning on a Dell 690 computer. Here is a copy/paste from the release notes

''Dell 690 - Image saving and copying is no longer allowed on the 690 computers.
Run Parameters will not allow saving any images if running on a 690 computer.''
''Dell T7500 – Recommend only saving up to 10% of images on the T7500. Saving more than 10% will degrade run performance. If more than 10% of images are saved the application will display the warning message shown below. The user can proceed despite the warning and start the run.''

There are lots of improvement in this software upgrade.

[mattanswers]

We always ask for the image files whenever we get sequencing done. This is raw data and as such it is usually considered good scientific practice to hold on to. Also, as base calling software gets better, and it does at a swift pace, it is possible that the image files could be used again. We have a tape drive so that we can save image files to tape.

[agagne]

We regularly use the images for our analysis, as we've found that many of our users get significant improvements from adjusting cluster identification cycles.

We've retained images for over a year and are going to stop this practice soon. We've only had a couple request from users to retrieve this type of data.

[hoisinjl]

We save images for a month and routinely need to save images to use specific cycles as a control lane or to adjust the cluster identification cycles. According to the RTA 1.8 release notes there will still be the ability to save images (recommded 10%, but only if you have a Dell T7500, which I don't have) and there is still an OLB option. I've asked my FAS to help get a straight answer as this affects many of our practices!!!

[simonandrews]

For many of our application we need to reprocess from images as the existing SCS makes a mess of cluster calling in heavily biased sequences. If the new software has improved options for cluster identification we might be OK, but I'm not sure how you'd test that without risking losing a whole run.

[hoisinjl]

So my FAS has informed me that OLB 1.8 will be available, hopefully later this week. Once that is installed on the computer cluster, we'll be able to test the new limitations. If you want to take advantage of the new v5 chemistry, however, you will need to upgrade the software. The only way to still save images (and I'm not sure on the impact on the run this will have) is to either currently have or upgrade to the Dell T7500.

[GW_OK]

From my FAS everything should be available now. Cluster identification should be better, as you can now go to 7-800k/mm2, which is ~378k-432k cpt. I figure you might be able to do a whole human genome on a single GAIIX if you do PE151. My IT guy is none too happy about the loss of qseqs, though.

Edit: Yup. everything up on the ftp site.

[hoisinjl]

Have you tried the BCl Converter? It will convert the .cbl files from BaseCalls to qseq files.

[GW_OK]

Yeah, but it's still an extra step.

[james hadfield]

I know the replies do not suggest it but I would think most users are perfectly happy to dispense with saving images. It is only for a minority of runs that keeping the images is useful and I for one am very happy to see them gone the way of the Dodo.

I would be more interested in seeing the intensitites do the same thing ASAP so we can more eaily store and move data around.

[csquared]

There is also going to a price increase on the new v5 reagents, although how much has not been decided/announced.

You can still save intensities, just not the images. This change is related to compute capabilities relative to data production rate, just like on the HiSeq. If you don't have a newer Dell (T7500), I would be very cautious about upgrading. The older machines don't handle the new conditions well at all.

The extra step in the conversion to qseq files is a bit of a pain, but temporary (I hope). Consider it a warm up for the HiSeq. All current HiSeq users have to go through the extra conversion step as the conversion is not a part of the Illumina pipeline at this time. Hopefully soon?

[flxlex]

He did relate one rather surprising detail about this software; it will not allow saving of images at all! (other than thumbnails). This means no more possibility to do offline base calling

I just heard that the HiSeq does not save images either...

[GW_OK]

The HiSeq doesn't image the same way as the GA2. I think it's more like a scanner than a camera?

[suludana]

What about the runs without PhiX but with the need of a posterior normalization (using PhiX from a previus run)? In this case we need the images, isn't it?