Synapse Curator Transition Guide

What to Expect For This Brand New Feature

Synapse Curator is actively being developed, and your experience directly shapes how it improves. Here's what that means in practice:

• Things may change between sessions — features and interface details are still being refined.

• You may encounter bugs or rough edges, including the known issues described below.

• Feedback is genuinely valuable at this stage — if something feels confusing or buggy outside of what’s described below, please report it to us at our NF-OSI Help Desk.

We appreciate your patience and want to hear both the issues you run to as well as what seems to help you most.

What's Better in Synapse Curator

Here's what you'll notice right away, compared to Data Curator App:

The right metadata template is recommended for you. Instead of a “Select Template” step and having to figure out which template applies to your dataset, in Synapse Curator one has already been suggested for you. We use your data sharing plan and any currently uploaded files to pre-create curation Tasks with the recommended template. Note: this is also a beta AI-based feature, so reach out if you have suggestions for improvement.

Your template loads much more quickly and right in Synapse. You still get a familiar, spreadsheet-like grid interface, but it is generated in a fraction of a time compared to Data Curator App. You also don’t have to navigate to an outside Google-based domain, which was an issue for a minority of institutions with strict blocking rules.

There's now an AI assistant to help you fill in your metadata. Once you're working in a template, the AI assistant can help auto-fill fields you point out, suggest values, and answer questions about what’s being asked for. It's there to reduce guesswork, especially for fields that aren't self-explanatory.

Validating and submitting your metadata is much simpler and faster. The Data Curator App workflow requires downloading a CSV and re-uploading it to validate. You can now fill in your metadata and submit it (“Sync Changes”) directly in Synapse with overall fewer steps and hundreds of times faster validation.

Drop-down options are more relevant to the data. Previously, drop-down menus could show options that didn't make sense for your assay data — for example, seeing sequencing platforms for platform listed when annotating imaging data. The new infrastructure has enabled us to make a long-wished for improvement to conditionally show only the relevant subset. This is less confusing and reduces the chance of inconsistencies.

Finding the right value for certain fields is easier. Fields like modelSystemName now sync with our tools database for selectable values that also work with autocomplete, helping you use the recommended nomenclature. Note: if you don’t find the right system name, you can submit a custom free text value for now (see below).

You can use a new/provisional term without emailing us first. If you can't find the right controlled value for a field and need to add a term, you can still submit (though the UI will still highlight it as invalid). This is unlike in Data Curator App, where you are completely blocked until one new term is added. The AI assistant is there to facilitate this flexibility, so it is recommended to use it first — it may find a match you missed or suggest how to add a new term with the highest acceptance probability. We want to reduce the back-and-forth for cases where something genuinely isn't in our metadata dictionary yet. We will still follow-up as needed after detecting new terms. Note: we’ll assess the merit of this new process to see whether it works well for everyone.

Getting Started: Step by Step

1	Open the new “Tasks and Actions” tab in your project page.
2	Find your Task and click to open the Curator. You'll see a list of tasks for your dataset folders. Find the task name that exactly matches the data folder you want to annotate or, in some cases, matches the closest parent folder. Tip: Leverage your browser’s control / CMD+F to find the name if you have a large list. Then click "Open Curator" under Actions, which launches the app in a new tab. Each link is specific to that dataset — think of it as a direct launcher for that folder's metadata.
3	Fill out your metadata. You'll land in the curation grid where you can enter metadata for your files. A few tips: Use drag-to-fill (hold Shift key and drag) and copy-and-paste (Ctrl+C and Ctrl+P) as usual to populate fields quickly Click "Open Chat" at any time to ask the AI assistant for help Each action is automatically saved, so you can use Undo and Redo.
4	“Sync Changes” to apply the annotations — equivalent to "Submit" in the Data Curator App, but you can do it iteratively.

A Few Things to Know

Synapse Curator is new, so there are a couple of rough spots we're still smoothing out, including some things that work differently from Data Curator App:

“Sync Changes” often to submit your current values as annotations, and try to complete all annotations in the same session or leave the tab open if you need to make more changes. Though you can close the session tab and open it again later using the same “Open Curator” link, the reloading experience is not yet optimized.

When “Open Curator” link reopens a session with some changes already, you will see the app “replay” your previous actions. This is the app catching up on your session history. You don't need to do anything, but you may need to wait a couple of minutes for it to finish.

The AI assistant can see the current validation errors and the template schema and do its best to provide valid suggestions, but be aware that AI has limitations.

Hover over column headers for descriptions of expected values. If you're unsure what a field is asking for, hovering over the column header, try asking the AI assistant, or look at the available drop-down options for clues.

The best workflow is one person responsible for curation of each dataset. Unlike the DCA-generated spreadsheets that can be editable by anyone you share the link with, Synapse Curator currently creates a private grid session that cannot be shared (yet). You will be able to invite others to your session in a much later update.

Walk-through video

This walk-through video demonstrates the step-by-step instructions and the AI assistant.

Frequently Asked Questions (FAQ)

• Do I have to switch to Synapse Curator right away? No. The previous Data Curator App is still available during this transition period. You can ignore the Synapse Curator tasks set up if you don’t want to use this yet.

• Is it OK if, for the same dataset, I contribute metadata through (a) Data Curator App and then later use Synapse Curator? What about the (b) vice versa scenario? Scenario (a) is expected and should present no issue; you can open up any dataset that has already been annotated with Data Curator App and view/update the metadata in Synapse Curator (as long as the curation task has been setup for that dataset). We will also support scenario (b) — you should still be able to find your dataset in the Data Curator App and go through the steps there — but there may some unexpected cases we cannot account for yet in this transition period.

• Will I see a curation task for a new folder I add later? Not automatically — curation tasks aren't created for new folders yet. If you add a new data folder and need a task for it, just let us know and we'll set it up. In a future update, you'll be able to create tasks yourself directly with UI in Synapse.

• How do I edit or delete a curation task? You can't manage tasks directly yet — self-service task management UI is coming in a future update. In the meantime, reach out to us if you need a task removed or changed.

• The suggested template doesn't look right for my dataset. What should I do? Reach out to a data manager you are already in contact with or create an issue at our NF-OSI Help Desk .

• Does Synapse Curator work if I have thousands of files in a folder? Currently, this approaches the limit of what makes sense in the current UI. Contact us to discuss how we can help you download a CSV template to fill programmatically or splitting the dataset into subfolders (batches) and creating a curation task for each.

• I can't find the term I need in the drop-down. What are my options? First, try asking the AI assistant; it may be able to suggest a term that already exists in our list that matches what you're looking for. If nothing fits, you can submit a provisional new value directly in the app. We'll follow up if any changes are needed on our end.

• Who do I contact if I run into a problem? Respond to your notification email or create a new ticket at our NF-OSI Help Desk. We're actively monitoring feedback during this early period and want to hear about any issues you encounter.