Ryans help for CFG data import

[ReneRanzinger]

Based on the collection in #91 from @senaarpinar and @akulmehta we send an email to Ryan on 7/31/2023.

[ReneRanzinger]

Hi Ryan,

I hope you are doing well. It’s going to be a long email. Sorry about this. Together with Akul and Jamie, we are still working on importing the old CFG data into the glycan array data repository. There were a number of questions and issues that came up in the process, and we hope that you will be able to help with them. The questions/issues are related to two main topics

1. MIRAGE metadata for the CFG data

2. Glycan sequence issues in datasets

3. MIRAGE metadata
   We realized that it may be impossible to get the MIRAGE metadata for all old datasets. Some of the metadata is on the CFG webpage but many things are missing. You will probably not be able to help with information related to scanning, image processing and data processing but we hope that you help fill the gaps related to slide, spot, features and printing. We are hoping that you can give us one default set of metadata that we can just use for all old datasets. Based on the excel spreadsheets that we worked on together over the last years we made template for the information we would like to have from you:
   https://docs.google.com/spreadsheets/d/1rGd9g3WKs0E8bflBXGZfY1N7zYJN0Vk0RovSAZ4jk34/edit#gid=1294057993

We hope that you will be able to fill column J (Value) for the different sheets:

• “Slide metadata”: Information about the slide.
• “Printer Metadata” and “Printrun Metadata”: information about the printer and the printing.
• “Feature metadata”: Information for each glycan probe on the slide. It's probably not possible to get these for each single glycan. But maybe we can use a default information set for all glycans. If not we will have to leave this empty.
• “Spot metadata”: A default set of information that we can apply to all spots.
• We would also need a value for the glycan concentration on the spot. As a placeholder we are using 100uM right now.
  For rows that require a unit of measurement (column I) please append the correct unit in column J.

As said above we will not be able to get exact metadata for each single printed slide (or even slide version), spot or glycan. But if you can provide us with one “representative” set of information we can apply them for each slide/spot/glycan and use the general comment field to let users know, that this is representative information and that the actual information may have been different.

2. Glycan sequence issues in datasets
   While trying to import the glycan array data we noticed several issues with the provided masterlists (GAL files) and processed data files.

a) We have gone over all the masterlists and compiled a list of glycan structures with all their different sequence representations and tried to come up with a single master sequence for each glycan. The assumption is that for example “561 Sp0” is always the same glycan (+linker) no matter in which array version it is printed. The following spreadsheet contains the 17 glycans that we could not resolve, i.e, could not decide the correct sequence.

https://docs.google.com/spreadsheets/d/1MbIuORIjRr4l02pJxRV_q0P3E8DMnMxP1fAEqu8L_lk/edit#gid=0

In column A you find the name/identifier (e.g. 327 Sp10) of the Glycan. In column B you find the sequence associated with this and in column E in which CFG version this sequence-name/identifier combination was used. For example row 4+5: 561 SP0 : In 5.4 the sequence has a GalNAc in the middle, in the older versions a GlcNAc. We are not able to figure out what the right sequence for 561 SP0 is. It may be an error that was found in 5.4 and corrected or it's a typo in 5.4 and all other versions are correct. There are some comments Sena/Akul left in column A or B describing the problem/reason for not being able to assign the correct sequence for these glycans.

We hope you will be able to help identify the correct sequences for these 17 glycans.

b) As a second step, we went over all the processed data files (for all versions) and tried to match the sequence given in the Excel files to the glycans we gathered from the masterlists. We have checked all synonyms (all different representations in that version’s masterlist) and the master sequences to find the matching glycan (and linker).

The following spreadsheet contains a list of sequences from all CFG processed data files that we could not match with certainty with the structures given in masterlists:

https://docs.google.com/spreadsheets/d/1EUklE9aVLcPPmfTesvunIB9DLa_DVZ2kmHYu9pnEYE8/edit#gid=0

Column A shows the sequence that was used in the data file and that we could not match to any sequence from the masterlists. Column B shows the number of experiments/files that use that sequence. Column C the versions that this sequence was used in. In E you see the glycan name that we (based on the GAL file) assume the glycan is. And in column F the sequence from the masterlist associated with the glycan name from column E.

You can also mouse over cells in column A. Sena left for most of them comments to highlight the problem we see.

We hope that you will be able to help resolve the questions.

If you need more information or want to talk about your findings, please let us know. We are happy to set up an audio/Zoom meeting with you and discuss these issues. If you think we should also involve somebody else in this discussion please let us know.

Your help in this matter would be really appreciated, since we want to make the CFG import as complete and as correct as possible. Without your help this will not be possible.

All the Best,
René