|
 
 |
| EDITORIAL |
|
| Year : 2021 | Volume
: 12
| Issue : 1 | Page : 1-3 |
|
Re-Imagining Machine Learning in Dental Research: A Lesson Learned from the COVID-19 Pandemic
Jafar Kolahi1, Mohammad Seyedhamzeh2, David G Dunning3, Nader Kalbasi4
1 Independent Research Scientist, Associate Editor of Dental Hypotheses, Isfahan, Iran
2 Independent Research Scientist, Zanjan, Iran
3 Department of Oral Biology, College of Dentistry, University of Nebraska Medical Center, Lincoln, Nebraska, USA
4 Department of Oral and Maxillofacial Pathology, School of Dentistry, Isfahan(Khorasghan) Branch, lslamic Azad University, Isfahan, Iran
| Date of Submission | 10-Dec-2020 |
| Date of Decision | 10-Jan-2021 |
| Date of Acceptance | 15-Jan-2021 |
| Date of Web Publication | 2-Mar-2021 |
Correspondence Address:
Jafar Kolahi
No. 24, Faree 15, Pardis, Shahin Shahr, Isfahan
Iran
 Source of Support: None, Conflict of Interest: None  | Check |
DOI: 10.4103/denthyp.denthyp_169_20
How to cite this article:
Kolahi J, Seyedhamzeh M, Dunning DG, Kalbasi N. Re-Imagining Machine Learning in Dental Research: A Lesson Learned from the COVID-19 Pandemic. Dent Hypotheses 2021;12:1-3 |
How to cite this URL:
Kolahi J, Seyedhamzeh M, Dunning DG, Kalbasi N. Re-Imagining Machine Learning in Dental Research: A Lesson Learned from the COVID-19 Pandemic. Dent Hypotheses [serial online] 2021 [cited 2023 Jun 5];12:1-3. Available from: http://www.dentalhypotheses.com/text.asp?2021/12/1/1/310535 |
At this time, humankind is experiencing on a global scale a frightening infectious disease caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). As of September 25, 2020, more than 32.3 million cases have been reported in 188 countries, resulting in more than 984,000 official reported deaths.[1],[2] During the course of human history, there have been several pandemics, for example, plague and smallpox. However, one of the main differences between COVID-19 and previous pandemics is the availability of computer-based technologies. Paralleling the global growth of COVID-19, studies using computer-based technologies have emerged (e.g., machine learning) to fight against SARS-CoV-2.[3],[4]
The term “machine learning” is a subset of artificial intelligence in which a mathematical model is built using sample data, known as “training data,” in order to make forecasts without being obviously programmed to do so. Machine learning algorithms easily identify trends and patterns among multidimensional and multi-variety data without any human intervention. Importantly, machine learning algorithms must be chosen carefully for a specific purpose and require massive data sets of high quality on which to train.
A good example for utilizing machine learning in biomedical research would be “Chemprop — Machine Learning for Molecular Property Prediction.”[5] In this project, Massachusetts Institute of Technology researchers used an artificial neural network and deep learning approach to predict the probability that a molecule will inhibit the SARS-CoV-2 main protease. [Figure 1] depicts a structural view of the SARS-CoV-2 main protease, also called 3C-like protease, and its interaction with a potential antiviral drug, Boceprevir. Boceprevir is a protease-inhibitor used successfully to treat hepatitis caused by the hepatitis C virus, created by way of molecular docking using the Molegro Virtual Docker (Molexus IVS, Rørth, Denmark).[6] | Figure 1 Secondary structure view of the SARS-CoV-2 main protease (3C-like protease) in complex with potential inhibitor Boceprevir (colored by green and labeled as U5G_503 [A] (85)).
Click here to view |
However, correct use of the Chemprop structure of molecules, Boceprevir (C27H45N5O5), requires a conversion to a SMILES format as follows: “CC1([C@@H]2[[email protected]]1[[email protected]](N(C2)C(=O)[[email protected]](C(C)(C)C)NC(=O)NC(C)(C)C)C(=O)NC(CC3CCC3)C(=O)C(=O)N)C” that, in turn, could be used for machine learning. The activity score predicated by Chemprop for Boceprevir was 9.26, indicating it is inactive against SARS-CoV-2 main protease. Also, Massachusetts Institute of Technology researchers translated structural information of the spike protein of SARS-CoV-2 into music using machine learning.[7]
Nevertheless, a PubMed search with query “machine learning” and filters yielded the following results: dental journals at Oct 16, 2020, revealed only 97 results, see [Figure 2], the Journal of Craniofacial Surgery published the most number of articles (11.1%), followed by the Journal of Dental Research (8%), and then, Oral Surgery, Oral Medicine, Oral Pathology and Oral Radiology (7%). The articles entitled “Predicting oral malodour based on the microbiota in saliva samples using a deep learning approach”[8] received the most online attention (Altmetric score: 17). Among 65 articles that received online attention (Altmetric score ≥ 1), there was no correlation between Altmetric score and number of citations (source of citations data: Dimensions, Pearson’s r = 0.025, 95% confidence level: −0.207 to 0.256, and P = 0.832). | Figure 2 The number of articles published in dental journals regarding machine learning. Also, the number of articles during the next 5 years forecasted using an exponential smoothing algorithm (upper part). Author keyword co-occurrence network analysis of machine learning articles (lower part).
Click here to view |
However, the broader dental research community, including clinical practitioners, could pay more attention to machine learning as a futuristic concept for dental diagnosis,[9] prediction of root caries,[10] oncology,[11],[12] orthodontics,[13] radiographic detection of apical lesions,[14] caries detection,[15] dental altmetrics,[16] and so on.
Recently, a deep learning model (available via Chemprop[5]) was used to successfully predict a new antibiotic to inhibit the growth of Escherichia More Details coli known as Halicin, structurally divergent from conventional antibiotics and exhibiting bactericidal activity against a wide range of well-known pathogens in vivo.[17] This same methodology could be performed rapidly and inexpensively to predict new chemicals against druggable proteins of well-known dental pathogens such as Streptococcus mutans[18] and Enterococcus faecalis.[19]
Finally and critically, dental journal editors, grant providers, research managers, educational planners, and clinical professionals could direct more attention to artificial intelligence and machine learning as a rapid, effective, and inexpensive tool to solve complicated clinical problems.[20]
Financial support and sponsorship
Nil.
Conflicts of interest
The authors reported no conflicts of interest.
| References | |  |
| 1. | Dong E, Du H, Gardner L. An interactive web-based dashboard to track COVID-19 in real time. Lancet Infect Dis 2020;20:533-4.  |
| 2. | |
| 3. | Bachtiger P, Peters NS, Walsh SL. Machine learning for COVID-19—asking the right questions. Lancet Digit Heal 2020;2:e391-2. |
| 4. | Lalmuanawma S, Hussain J, Chhakchhuak L. Applications of machine learning and artificial intelligence for Covid-19 (SARS-CoV-2) pandemic: a review. Chaos Solit Fract 2020;139:110059. |
| 5. | |
| 6. | RCSB PDB − 6ZRU: Crystal structure of SARS CoV2 main protease in complex with inhibitor Boceprevir. Available from: https://www.rcsb.org/structure/6ZRU. Accessed September 25, 2020. |
| 7. | |
| 8. | Nakano Y, Suzuki N, Kuwata F. Predicting oral malodour based on the microbiota in saliva samples using a deep learning approach. BMC Oral Health 2018;18:128. |
| 9. | Mupparapu M, Wu CW, Chen YC. Artificial intelligence, machine learning, neural networks, and deep learning: Futuristic concepts for new dental diagnosis. Quintessence Int 2018;49:687-8. |
| 10. | Hung M, Voss MW, Rosales MN et al. Application of machine learning for diagnostic prediction of root caries. Gerodontology 2019;36:395-404. |
| 11. | Kumar S, Awan KH, Patil S, Bharkavi SKI, Raj AT. Potential role of machine learning in oncology. J Contemp Dent Pract 2019;20:529-30. |
| 12. | Sultan AS, Elgharib MA, Tavares T, Jessri M, Basile JR. The use of artificial intelligence, machine learning and deep learning in oncologic histopathology. J Oral Pathol Med 2020;49:849-56. |
| 13. | Chen S, Wang L, Li G et al. Machine learning in orthodontics: Introducing a 3D auto-segmentation and auto-landmark finder of CBCT images to assess maxillary constriction in unilateral impacted canine patients. Angle Orthod 2020;90:77-84 |
| 14. | Ekert T, Krois J, Meinhold L et al. Deep learning for the radiographic detection of apical lesions. J Endod 2019;45:917-22.e5. |
| 15. | Casalegno F, Newton T, Daher R et al. Caries detection with near-infrared transillumination using deep learning. J Dent Res 2019;98:1227-33. |
| 16. | Kolahi J, Khazaei S. Altmetric analysis of contemporary dental literature. BDJ 2018;225:68-72. |
| 17. | Stokes JM, Yang K, Swanson K et al. A deep learning approach to antibiotic discovery. Cell 2020;180:688-702.e13. |
| 18. | Horst JA, Pieper U, Sali A et al. Strategic protein target analysis for developing drugs to stop dental caries. Adv Dent Res 2012;24:86-93. |
| 19. | Cathro P, Mccarthy P, Hoffmann P, Zilm P. Isolation and identification of Enterococcus faecalis membrane proteins using membrane shaving, 1D SDS/PAGE, and mass spectrometry. FEBS Open Bio 2016;6:586-93. |
| 20. | Schwendicke F, Samek W, Krois J. Artificial intelligence in dentistry: chances and challenges. J Dent Res 2020;99:769-74. |
[Figure 1], [Figure 2]
|
Search Pubmed for
