SiDock@home

From BOINC Projects
Revision as of 12:04, 23 May 2026 by Al Piskun (talk | contribs) (update and infobox)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search




SiDock@home
Project
StatusActive
CategoryBiochemistry / Drug Discovery
ComputeCPU
RequiresMicrosoft Visual C++ Redistributable 2019 (Windows only)
Development
DeveloperNatalia Nikitina, Maxim Manzyuk, Marko Jukić, Črtomir Podlipnik, and team
AuthorNatalia Nikitina, Maxim Manzyuk
SponsorKarelian Research Center of the Russian Academy of Sciences; Slovenian Research Agency (ARRS); Slovenian Ministry of Science and Education (HPC-RIVR); COVID.SI project
Initial releaseDecember 17, 2020  (6 years ago)
Repositoryhttps://gitlab.com/Jukic/cmdock (CmDock)
Software
Written inC++
Operating systemWindows, Linux, macOS (x86-64); ARM (in development)
BOINC statistics
Stats as ofMarch 2022
Performance56 TFLOPS
Total users7,540
Active hosts~8,000
Metadata
Websitehttps://www.sidock.si/sidock/
SiDock@home logo

SiDock@home is an international volunteer distributed computing project that uses BOINC to perform high-throughput virtual drug screening against dangerous viral pathogens. Launched in December 2020 as a direct outgrowth of the citizen-science effort COVID.SI, SiDock@home harnesses the idle processing power of volunteers' personal computers worldwide to computationally sift through enormous chemical libraries in search of molecules capable of disabling key viral proteins — work that would otherwise require dedicated supercomputer time costing millions of dollars.

Why SiDock@home?

The challenge of discovering new antiviral drugs is staggering in scale. The space of all possible drug-like molecules is estimated at roughly 1080 distinct structures — larger than the number of atoms in the observable universe.[1] Even if researchers restrict their attention to the "small molecule universe" — synthetically feasible organic molecules with a molecular weight under 500 daltons — the pool still exceeds 1060 candidates.[2] Evaluating each candidate experimentally is completely impractical; computational pre-filtering via molecular docking allows researchers to reduce that ocean of possibilities to a manageable shortlist for laboratory follow-up.

Traditional high-performance computing (HPC) resources are expensive, heavily scheduled, and not always accessible to academic research groups. Volunteer computing using the BOINC platform offers a highly scalable and cost-effective complement: instead of one supercomputer, researchers recruit tens of thousands of ordinary personal computers donating their spare cycles. SiDock@home was created to tap exactly that potential, directing it toward a medically urgent goal — finding small molecules that could stop SARS-CoV-2 and other dangerous pathogens in their tracks.[3]

Colorized scanning electron microscope image of SARS-CoV-2 (orange) emerging from the surface of a cell. SiDock@home's first mission targeted multiple proteins essential to this virus's life cycle.

Goal

The core objective of SiDock@home is to perform high-throughput virtual screening (HTVS): automatically docking millions of candidate compounds against carefully chosen protein targets from dangerous viruses, then returning ranked lists of the most promising "hits" for the research team to investigate further.[4]

The project's initial mission was to screen a specially developed compound library against a set of SARS-CoV-2 protein targets playing critical roles in the viral life cycle, including the main protease (3CLpro), papain-like protease, and envelope protein. Beginning in 2024, the community voted to expand the project's scope to the Ebolavirus glycoprotein (GP1), broadening SiDock@home into a general-purpose drug discovery platform.[5]

The central computational challenge involves predicting how well a small molecule (the ligand) fits into the active site of a target protein — analogous to finding a precisely shaped key for a very complex lock. As the project's own description puts it, the task is like finding a needle in a haystack, because the number of possible molecular structures is so astronomically large.[6]

History

SiDock@home grew directly out of COVID.SI, a Slovenian citizen-science project launched in 2020 to fight SARS-CoV-2 using distributed computing.[7] COVID.SI developed the initial docking software and compound library; SiDock@home was created to scale that effort up dramatically by plugging into the established global BOINC volunteer computing network.

A public test phase began in November 2020, running on the subdomain fightcovid.boinc.ru/sidocktest/.[8] On 17 December 2020, the project was declared production-ready and moved to its permanent home at sidock.si/sidock/, with all test-phase credits and user accounts migrated.[9] BOINC project administrator David Anderson welcomed SiDock@home to the BOINC portal on 23 December 2020, noting its sponsorship by COVID.SI and the Karelian Research Center of the Russian Academy of Sciences.[10]

Within just a few months of its public launch, SiDock@home had grown to the scale of a modern supercomputer and subsequently became an independent general drug discovery project, with SARS-CoV-2 as its founding mission.[4]

The first compound library of one billion molecules was prepared with donated computational resources from Microsoft Azure.[11]

Methods

Molecular Docking

The scientific core of SiDock@home is molecular docking. In this computational technique, a software program predicts the preferred orientation of a small candidate molecule (the ligand) when it binds to a target protein, and assigns the result a docking score that approximates the binding affinity. Molecules with the best scores are prioritized for further wet-lab testing.

The ideal ligand must be complementary in both shape and chemical properties to the binding site of the target biomolecule.[12] This complementarity is only one of many prerequisites for a viable drug candidate, but it provides an efficient first filter: by screening millions of compounds computationally before any laboratory work begins, researchers can concentrate expensive and time-consuming experimental resources on the most promising hits.

CmDock / CurieMarieDock

SiDock@home uses a custom molecular docking application called CmDock (also known as CurieMarieDock), developed by the project team as a fork and optimization of RxDock (itself derived from the open-source rDock program).[13] CmDock is faster than its predecessors and is available for Windows, Linux, macOS (x86-64), with an ARM version in development. Its source code is hosted publicly at GitLab. The use of a purpose-built, open-source docking engine — rather than a commercial program — allows the team to probe a unique region of chemical space and generate scientific knowledge not available from prior studies.[4]

How It Works

When a volunteer's computer connects to SiDock@home through the BOINC client, it receives a work unit: a small batch of candidate compounds paired with the 3D structure of the current target protein. The local CmDock application then systematically places and scores each compound inside the protein's binding site. Completed results are uploaded to the project server, where they are validated and incorporated into the running dataset. Because many volunteers contribute simultaneously, the project can screen enormous libraries in a fraction of the time a single research group could manage alone.

Ribbon diagram of a protein structure. SiDock@home volunteers computationally dock millions of candidate ligands into binding pockets of viral proteins like this to identify potential drug candidates.

Why BOINC?

BOINC (Berkeley Open Infrastructure for Network Computing) is the world's leading platform for volunteer distributed computing, providing proven infrastructure for task distribution, result validation (via redundant replication), and credit accounting. Using BOINC allowed SiDock@home to tap an already-established global community of volunteers without needing to build scheduling and client infrastructure from scratch. The platform's heterogeneous "master-worker" model is especially well suited to virtual screening, where each docking calculation is independent and can be performed on any capable CPU without inter-node communication.[14]

By harnessing idle CPU cycles on thousands of ordinary personal computers worldwide, SiDock@home achieves a level of computational throughput comparable to a dedicated supercomputer — making large-scale drug discovery accessible to research groups that could never afford dedicated HPC time.

Research Targets

Targets are selected by the scientific team and, in a community-oriented spirit, sometimes put to a volunteer vote. Targets screened or announced to date include:

  • Multiple SARS-CoV-2 proteins, including the main protease (3CLpro) with both active and allosteric sites, the papain-like protease (PLpro), and the envelope protein (a viroporin conserved across viral variants)[15]
  • Ebolavirus glycoprotein GP1 (Target #23, introduced in July 2024, chosen by community vote with 174 out of 425 votes)[16]

As of the test phase in late 2020, the team had identified 59 potential SARS-CoV-2 targets for future investigation.[17]

Project Team and Sponsors

SiDock@home is a genuinely international collaboration spanning Russia, Slovenia, and Poland.

  • Natalia Nikitina (also published as Natalia Kukushkina) — co-founder and lead developer; Institute of Applied Mathematical Research, Karelian Research Center of the Russian Academy of Sciences, Petrozavodsk, Russia[18]
  • Maxim Manzyuk — co-founder and server administrator; Federal Research Center "Computer Science and Control" of the Russian Academy of Sciences, Moscow (also associated with the BOINC.RU internet portal)[19]
  • Marko Jukić — biochemistry lead; Faculty of Chemistry and Chemical Engineering, University of Maribor and Faculty of Mathematics, Natural Sciences and Information Technologies, University of Primorska, Slovenia. Developer of CmDock.[20]
  • Črtomir Podlipnik — computational chemistry; Faculty of Chemistry and Chemical Technology, University of Ljubljana, Slovenia[21]

Funding and infrastructure support has been provided by:

  • The Scholarship of the President of the Russian Federation for young scientists and graduate students (project SP-609.2021.5)
  • The Slovenian Research Agency (ARRS), programmes P2-0046, J1-2471, and P1-0201
  • The Slovenian Ministry of Science and Education, infrastructure grant HPC-RIVR and programme grant OP20.04342
  • Microsoft, which donated Azure cloud computing resources to prepare the initial billion-compound screening library[22]

Statistics and Performance

By the end of March 2022, SiDock@home had 7,540 registered users contributing roughly 8,000 active computers, delivering a combined processing power of approximately 56 teraFLOPS as measured by BOINC.[23] That figure is consistent with the computational footprint of a significant modern supercomputer.

Registration for new accounts uses a Crunch_4Science invitation code (not required when joining via BOINC Manager directly).[24]

Scientific Results

The following peer-reviewed publications have been produced by the SiDock@home team:

  1. Nikitina, Natalia and Evgeny Ivashko. Optimization of the Workflow in a BOINC-Based Desktop Grid for Virtual Drug Screening. (2022). DOI: 10.1007/978-3-031-22941-1_50.
  2. Jukić, Marko, Sebastjan Kralj, Natalia Nikitina and Urban Bren. Bioinformatic and MD Analysis of N501Y SARS-CoV-2 (UK) Variant. (2021). DOI: 10.1007/978-3-030-86582-5_1.
  3. Nikitina, Natalia, Maxim Manzyuk, Črtomir Podlipnik and Marko Jukić. Volunteer Computing Project SiDock@home for Virtual Drug Screening Against SARS-CoV-2. (2021). DOI: 10.1007/978-3-030-86582-5_3.
  4. Nikitina, Natalia, Maxim Manzyuk, Črtomir Podlipnik and Marko Jukić. Performance Estimation of a BOINC-Based Desktop Grid for Large-Scale Molecular Docking. (2021). DOI: 10.1007/978-3-030-86359-3_26.
  5. Nikitina, Natalia, Maxim Manzyuk, Marko Jukić, Črtomir Podlipnik, Ilya Kurochkin and Alexander Albertian. Toward Crowdsourced Drug Discovery: Start-Up of the Volunteer Computing Project SiDock@home. (2021). DOI: 10.1007/978-3-030-92864-3_39.

See Also

References

  1. ↑ SiDock@home. BOINC Synergy. Retrieved 2026-05-23}.
  2. ↑ Ertl, Peter.(2003}).Cheminformatics Analysis of Organic Substituents. Journal of Chemical Information and Computer Science. pp. 374–380. DOI: 10.1021/ci0255782.
  3. ↑ "Volunteer Computing Project SiDock@home for Virtual Drug Screening Against SARS-CoV-2". In:(2021}).Computer Science Protecting Human Society Against Epidemics. Springer. DOI: 10.1007/978-3-030-86582-5_3.(IFIP Advances in Information and Communication Technology, vol. 616).
  4. ↑ 4.0 4.1 4.2 "Toward Crowdsourced Drug Discovery: Start-Up of the Volunteer Computing Project SiDock@home". In:(2021}).Supercomputing. RuSCDays 2021. Springer. DOI: 10.1007/978-3-030-92864-3_39.(Communications in Computer and Information Science, vol. 1510).
  5. ↑ BOINC Project SiDock@home. Gridcoinstats. Retrieved 2026-05-23}.
  6. ↑ SiDock@home. SiDock@home project. Retrieved 2026-05-23}.
  7. ↑ (2024-01-09}).Boinc Project: SiDock@home. Medium / Rotop100.com. Retrieved 2026-05-23}.
  8. ↑ SiDock@home [test]. Retrieved 2026-05-23}.
  9. ↑ SiDock@home Test Project — production announcement. Retrieved 2026-05-23}.
  10. ↑ (2020-12-23}).Welcome SIDock@home. BOINC. Retrieved 2026-05-23}.
  11. ↑ "Toward Crowdsourced Drug Discovery: Start-Up of the Volunteer Computing Project SiDock@home". In:(2021}).Supercomputing. RuSCDays 2021. Springer. DOI: 10.1007/978-3-030-92864-3_39.(CCIS, vol. 1510).
  12. ↑ SiDock@home. SiDock@home project. Retrieved 2026-05-23}.
  13. ↑ SiDock@home [test — New application: CurieMarieDock]. Retrieved 2026-05-23}.
  14. ↑ "Performance Estimation of a BOINC-Based Desktop Grid for Large-Scale Molecular Docking". In:(2021}).Parallel Computing Technologies. PaCT 2021. Springer. DOI: 10.1007/978-3-030-86359-3_26.(Lecture Notes in Computer Science, vol. 12942).
  15. ↑ BOINC Project SiDock@home. Gridcoinstats. Retrieved 2026-05-23}.
  16. ↑ BOINC Project SiDock@home. Gridcoinstats. Retrieved 2026-05-23}.
  17. ↑ SiDock@home [test project news]. Retrieved 2026-05-23}.
  18. ↑ Natalia Kukushkina — ResearchGate profile. Retrieved 2026-05-23}.
  19. ↑ "Volunteer Computing Project SiDock@home for Virtual Drug Screening Against SARS-CoV-2". In:(2021}).Computer Science Protecting Human Society Against Epidemics. Springer. DOI: 10.1007/978-3-030-86582-5_3.(IFIP AICT, vol. 616).
  20. ↑ "Volunteer Computing Project SiDock@home for Virtual Drug Screening Against SARS-CoV-2". In:(2021}).Computer Science Protecting Human Society Against Epidemics. Springer. DOI: 10.1007/978-3-030-86582-5_3.(IFIP AICT, vol. 616).
  21. ↑ "Volunteer Computing Project SiDock@home for Virtual Drug Screening Against SARS-CoV-2". In:(2021}).Computer Science Protecting Human Society Against Epidemics. Springer. DOI: 10.1007/978-3-030-86582-5_3.(IFIP AICT, vol. 616).
  22. ↑ "Toward Crowdsourced Drug Discovery". In:(2021}).Supercomputing. RuSCDays 2021. Springer. DOI: 10.1007/978-3-030-92864-3_39.(CCIS, vol. 1510).
  23. ↑ Nikitina, Natalia.(2022}).Lecture Notes in Computer Science. DOI: 10.1007/978-3-031-22941-1_50.
  24. ↑ SiDock@home. SiDock@home project. Retrieved 2026-05-23}.
Retrieved from "https://boincsynergy.ca/wiki/index.php?title=SiDock@home&oldid=1308"