[cosmarium]

Hi all,
Thank you very much for all your support. I put documentation together and I was able to have my samples repeated free of charge. Since I gave enough samples, they just repeated from what they had left. I am now analyzing my data
Thank you again,
C

[sklages]

Personally I do use the terms interchangeably for the company-provided indexes. Perhaps this is a bad habit. For customer-provided barcodes I called them exactly that -- customer-provided.

+1

That's how I deal with that as well :-)

[westerman]

The recent PDF I got from our Illumina rep was had columns "i5 index" (e.g., N501) and "i5 code" (e.g., TAGA...).

One of the FAQs from Illumina says:

Q: What level of indexing or barcoding is available?
A: Yes. The TruSeq Custom Amplicon Kit supports 96-plex indexing, TruSeq DNA Sample prep kits support 24-plex indexing, and Nextera kits support 96-plex indexing.

Another says:

Q: What level of multiplexing is supported on the GA with Nextera? When will this be available?
A: There will be 96 available barcodes using 12x8 indices, combining 12 Index 1(i7) with 8 Index 2 (i5) indices. Although kits are currently available, (the index kit supports multiplexing on all Illumina sequencing platforms) multiplexing on GA is not supported until updated SCS/RTA software is available.


So while Illumina does appear to use the word 'barcode' it is mostly uses the word 'index'.


Personally I do use the terms interchangeably for the company-provided indexes. Perhaps this is a bad habit. For customer-provided barcodes I called them exactly that -- customer-provided.

[GenoMax]

I agree. Perhaps it is better to think of them as below.

Index - Supplied in vendor kit(s)
Barcode - custom (can either be used in index read or integrated into sequence read).

I'm searching around and realizing that some people do use them interchangeably. Our core uses them as I stated, which I strongly prefer as it avoids confusion such as this.

[Heisman]

I'm searching around and realizing that some people do use them interchangeably. Our core uses them as I stated, which I strongly prefer as it avoids confusion such as this.

[sklages]

I have searched the web, and contacted Illumina tech. It seems that the two words are used fairly interchangeable. Thanks for your quick reply.
C

ACK, ... maybe context dependent .. I tend to talk about "barcode recognition" and I am asking about the "indices used" for sequencing :-)

[cosmarium]

I have searched the web, and contacted Illumina tech. It seems that the two words are used fairly interchangeable. Thanks for your quick reply.
C

[Heisman]

Where did you, or I can, find the official source of this information?

I'm sure there is at least one review out there that goes over indexing and barcoding. Perhaps somebody else can dig up a link. You can try searching around the internet for "indexing vs. barcoding" as well as it relates to sequencing.

[cosmarium]

Well I hate to be the potential bearer of bad news, but there is a large distinction between "indexes" and "barcodes". Indexes are placed between the flow cell adapter and the sequencing primer and require a completely separate read. Barcodes are placed at the end of the adapter sequence and will thus be the first bases read in a normal read.

Diagram:

Index
[flowcell adapter][index][sequencing primer][insert][sequencing primer][flowcell adapter]

Barcode
[flowcell adapter][sequencing primer][barcode][insert][sequencing primer][flowcell adapter]

If you used indexes and did not specify that, they may have not included an index read. If you used barcodes, then the barcode sequence should always appear at the end of one of the reads. Could you check a couple of your reads and see if the beginning/ends match any of your barcode sequences (or reverse complements)?

Where did you, or I can, find the official source of this information?

[GenoMax]

Like "westerman" said yesterday, there is an important lesson in all this.

Bottom line -these samples will need to be run again to get data you can actually use.

[cosmarium]

I asked them to rerun the casava with demutiplexing option, here is what they said:
" (...) this run was not run as a multiplexed run because you said the barcodes were in the sequence and not in the adapter. The files were run with configureBclToFastq.pl which would do a demultiplex if I had barcodes that were in the adapters & the associated barcode to sample list and had sequenced for an index. I then would have created a Sample sheet with this information to be used by the program to demultiplex.
I asked you several times about the barcoding of your samples and the mixing of your libraries.
However, you said the barcodes were in the sequence and NOT in adapters and therefore the sequencing was done without sequencing for indexes".

I did not said that, specially I don't call them barcodes. I call them index, I wrote in the submission form 12 index per lane, because the submission form (which they emailed to me) doesn't have a section to put whether multiplex or not. Aldo, I gave them a spreadsheet with these information per lane: index used, name of each of 12 samples, number of replicate, concentration; and I totalized concentration per lane.

I really wish we had stick to the other facility that we had been using. Never had a problem with them. These people we used now, were even discussing between themselves in front of me that I needed to provide cluster concentration, that I was not supposed to, etc. (a question I did in a forum before, 12 pM).

Thanks for your ideas and support. Comments are welcome...


C

[westerman]

As has been pointed out by "sklages", it is bad practice on the sequencing facility's part to just give you this data without de-multiplexing it.

It is also possible that *they* made an error and ran this as a non-multiplex flowcell. In that case, nothing can be done as far as demultiplexing the data you have.

You will need to get the facility to re-run the samples as multiplex.

Indeed if they ran it as a non-multiplex flowcell then it will need to be re-run. However there may be arguments on whose fault this is. The enclosed submission form only has a cryptic hand-written note at the bottom in the billing section about the indexing. Now it may be that the sequencing center needs to provide a better form. Or that cosmarium filled in the wrong form. In any case some one has made an expensive mistake from which, we hope, they learn something.

[GenoMax]

As has been pointed out by "sklages", it is bad practice on the sequencing facility's part to just give you this data without de-multiplexing it.

It is also possible that *they* made an error and ran this as a non-multiplex flowcell. In that case, nothing can be done as far as demultiplexing the data you have.

You will need to get the facility to re-run the samples as multiplex.

I did specify that they were multiplex index lanes. I also said I used TrueSeq kits.

I am attaching part of the submission form where it can be seen there that I even wrote it (but this is for curiosity...)

The point now is: the people from Maxwell computer cluster has not replied about the installation of CASAVA.

How else, can I try to search the index in a couple of sequences?

Thanks for your time,

C

[Heisman]

Indexes are done in a separate read, they won't just automatically appear in your normal read file, they have to be put somewhere. If they are not in the headers, they are probably at the beginning or end of the reads (only one of the two reads if you have paired end reads). Can you take a few reads, search for your indexes in them, and see where they are?

[cosmarium]

I did specify that they were multiplex index lanes. I also said I used TrueSeq kits.

I am attaching part of the submission form where it can be seen there that I even wrote it (but this is for curiosity...)

The point now is: the people from Maxwell computer cluster has not replied about the installation of CASAVA.

How else, can I try to search the index in a couple of sequences?

Thanks for your time,

C