Pathoplexus at one year: governance, data sharing and funding
Lee Katz, Nabil-Fareed Alikhan and guest host Clint return to Pathoplexus with Emma Hodcroft, Theo Sanderson and Arthur Shem Kasambula. Recorded on 22 August, ahead of the database’s first anniversary, the discussion covers community governance, sequence corrections, commercial use and early achievements. The guests explain why investment in sequencing also needs to support data sharing, and how redistribution can help prevent a database becoming a single point of failure.
Listen to Episode 145: Micro binfie Pathoplexus part 2 on SoundCloud
In this episode
Governance with member accountability
Emma Hodcroft explains that Pathoplexus was designed to serve the community generating and using pathogen genomic data. Its founders and members work with these data in genomics and public health, including informing government interventions. Community accountability was therefore part of the database’s design, rather than an administrative afterthought.
Pathoplexus uses a Swiss legal association structure called a Verein. Members can vote out the executive board and elect a replacement if they believe the organisation is moving in the wrong direction. Hodcroft notes that the board spans five continents. The guests also describe transparent decision-making, feedback from users and other experts, and documented, minuted meetings.
Open-source development and sequence corrections
Pathoplexus’s open-source software lets users report problems or request features through GitHub. They can also implement changes themselves and submit pull requests. One example is a request to select every sequence returned by a search, then deselect a small number. Sanderson invites listeners who want that behaviour to support the issue so it can be prioritised.
Sequence records have their own issue-reporting button, which at the time of recording creates a GitHub issue. A user might flag a collection date that does not fit when a lineage was circulating. Curators can contact the original submitters; for INSDC records whose submitters cannot be reached, they can add a curation to correct clearly erroneous data. Kasambula clarifies that corrections to directly submitted sequences remain the submitters’ responsibility: curators can flag an issue, but do not make the ultimate correction for them.
Commercial use and benefit sharing
The restricted-use terms limit use in publications and preprints, but do not themselves prohibit commercial use. Hodcroft describes the underlying equity problem: countries may share pathogen sequences that help companies develop vaccines, yet be unable to afford or obtain those vaccines. That outcome can undermine incentives to share data.
Pathoplexus does not claim to have solved pathogen access and benefit sharing. Instead, Hodcroft says its statutes commit it to comply when a globally accepted solution exists. The organisation wants to participate in a solution without deciding one on behalf of everyone involved.
She also says submitters do not explicitly give away rights over vaccine development from their sequences. In her account, this leaves open the possibility of a legal challenge, separate from and extending beyond the restricted-use period. She stresses that the law is complex and that a successful challenge is far from assured.
First-year milestones: mpox, Ebola and measles
Hodcroft reports that more mpox sequences had been submitted directly to Pathoplexus than directly to INSDC. Most of those Pathoplexus submissions were subsequently available in GenBank and ENA because Pathoplexus sent them there. Some submitters initially used restricted-use terms while preparing publications. She sees this as evidence that an accessible submission interface and temporary restrictions can encourage earlier sharing.
Sanderson describes a codebase of around 3,000 commits, with continuing improvements and support for more organisms. He says measles submission should be available by the time listeners hear the episode. Kasambula highlights the first Ebola sequence shared with Pathoplexus, crediting sequencing initiatives in Uganda and collaborators across the African continent.
Using Pathoplexus in teaching
The guests have seen Pathoplexus appearing in training workshops, tutorials and online teaching materials. Hodcroft describes two routes for learners: use the website to filter sequences by countries or mutations and download a dataset, or use the API to make requests and retrieve filtered data.
The website offers a gentler introduction before work such as phylogenetics, while API access supports more advanced workshops. Finding these teaching materials on GitHub or elsewhere online has been a welcome sign that people find the database useful.
Funding sharing, not just sequencing
Sanderson credits a development team extending beyond the guests, with contributors in Switzerland, the UK and elsewhere. Donations are welcome through the website’s funding page; one host notices that Bitcoin is an option. Sanderson contrasts the many millions of dollars spent on SARS-CoV-2 sequencing with the difficulty of attracting institutional support for effective sharing and analysis.
Hodcroft argues that investments in training, cheaper sequencing and wider access need to include what happens after data leave the sequencer. Otherwise, valuable data may remain private even when sharing could produce additional public-health benefits.
Redistribution is part of Pathoplexus’s approach to durability. Redistributors must carry the data-use terms and clearly identify restricted data. If Pathoplexus disappeared, another party could preserve access to the dataset under those rules. The aim is to avoid a single organisational failure causing thousands or millions of sequences to become unavailable.
Highlights
- 00:01:55 — The Swiss Verein structure and members’ power to replace the executive board
- 00:04:57 — Open-source development through GitHub issues and pull requests
- 00:06:55 — Reporting sequence errors and curating INSDC records
- 00:07:43 — Why directly submitted sequence corrections remain with submitters
- 00:08:11 — Commercial use, vaccine access and pathogen benefit sharing
- 00:12:04 — Direct mpox submissions as a first-year achievement
- 00:13:06 — Around 3,000 code commits, continuing improvements and measles support
- 00:13:55 — The first Ebola sequence shared with Pathoplexus
- 00:14:25 — Teaching with website filters, downloaded datasets and the API
- 00:15:49 — The wider development team, donations and institutional funding
- 00:17:18 — Funding data sharing alongside sequencing and preserving access through redistribution
In their own words
And for us, that was critical because we don’t want there to be one point of failure.
— Emma Hodcroft, 00:17:18
Who is talking
- Lee Katz (host)
- Nabil-Fareed Alikhan (host)
- Clint (host)
- Emma Hodcroft (guest)
- Theo Sanderson (guest)
- Arthur Shem Kasambula (guest)
Also mentioned: Andrew Page.
Tools and resources mentioned
Pathoplexus, INSDC, GenBank, ENA, GitHub, Excel.
Questions this episode answers
How is Pathoplexus governed?
Pathoplexus uses a Swiss association structure called a Verein. Members can vote out the executive board and elect a replacement, giving them a formal way to change the organisation’s direction.
How can users report incorrect Pathoplexus sequence metadata?
Sequence records have an issue-reporting button that, at the time of recording, creates a GitHub issue. Curators can contact submitters or add corrections to clearly erroneous INSDC records when submitters cannot be reached; directly submitted records remain the submitters’ responsibility to correct.
Do Pathoplexus restricted-use terms prevent commercial use?
No: the guests say these terms restrict publications and preprints, but do not themselves prohibit commercial use. Hodcroft also says submission does not explicitly give away rights over vaccine development, while cautioning that any legal challenge would be complex and not necessarily successful.
What did Pathoplexus achieve in its first year?
The guests report more direct mpox submissions than direct submissions to INSDC, with most subsequently passed to GenBank and ENA. They also highlight the first Ebola sequence shared with Pathoplexus, around 3,000 code commits, growing use in teaching and measles submission support expected by the episode’s release.
Could Pathoplexus data remain available if the database closed?
The guests explain that redistribution is permitted provided the data-use terms accompany the data and restricted records are clearly identified. This would allow another party to preserve access rather than making Pathoplexus a single point of failure.
This page was written from a machine transcript of the episode and checked against it. Listen to the episode for the whole conversation.
Show notes
Nabil and Lee bring a guest host Clint to talk with some of the people behind Pathoplexus http://pathoplexus.org/
- Dr. Emma Hodcroft - https://pathoplexus.org/about/eb
- Dr. Theo Sanderson - https://pathoplexus.org/about/development-team
- Mr. Arthur Shem Kasambula - https://pathoplexus.org/about/development-team