A Global Network Analysis of COVID-19 Vaccine Distribution to Predict Breakthrough Cases among the Vaccinated Population
Round 1
Reviewer 1 Report
The paper entitled "A Global Network Analysis of COVID-19 Vaccine Distribution to Predict Breakthrough Cases Among the Vaccinated Population" presents the research content and results from various aspects, which are novel and practical, but still have room for improvement in some aspects. The following are specific recommendations aimed at improving the quality and completeness of the paper:
1. The introduction section should more prominently highlight the challenges and impacts of resource scarcity. This will better emphasize the necessity and urgency of the research, thus establishing a stronger rationale for the study.
2. The analysis of existing research in the introduction is not sufficiently in-depth. There is a lack of systematic comparison between different methods and conclusions. It is recommended to include a detailed analysis of existing research, including a comparison of the advantages and disadvantages of various methods, and highlight the innovations in methodology or data application in this study to clarify its unique contributions.
3. The number of references is relatively low. It is suggested to increase the number of references from the relevant field to enhance the literature support and depth of the research background in the paper.
4. The quality of the figures needs improvement. It is recommended to enhance the clarity and professionalism of the figures, especially Figure 3, which is not very clear.
5. The description of the dataset used in the paper lacks detail. It is recommended to provide more detailed information in this regard.
6. The analysis of the performance of the deep learning models used in the paper is not thorough enough, especially concerning accuracy and efficiency. It is suggested to strengthen the discussion in this area.
7. It was suggested that a vision for future work be added.
The paper entitled "A Global Network Analysis of COVID-19 Vaccine Distribution to Predict Breakthrough Cases Among the Vaccinated Population" presents the research content and results from various aspects, which are novel and practical, but still have room for improvement in some aspects. The following are specific recommendations aimed at improving the quality and completeness of the paper:
1. The introduction section should more prominently highlight the challenges and impacts of resource scarcity. This will better emphasize the necessity and urgency of the research, thus establishing a stronger rationale for the study.
2. The analysis of existing research in the introduction is not sufficiently in-depth. There is a lack of systematic comparison between different methods and conclusions. It is recommended to include a detailed analysis of existing research, including a comparison of the advantages and disadvantages of various methods, and highlight the innovations in methodology or data application in this study to clarify its unique contributions.
3. The number of references is relatively low. It is suggested to increase the number of references from the relevant field to enhance the literature support and depth of the research background in the paper.
4. The quality of the figures needs improvement. It is recommended to enhance the clarity and professionalism of the figures, especially Figure 3, which is not very clear.
5. The description of the dataset used in the paper lacks detail. It is recommended to provide more detailed information in this regard.
6. The analysis of the performance of the deep learning models used in the paper is not thorough enough, especially concerning accuracy and efficiency. It is suggested to strengthen the discussion in this area.
7. It was suggested that a vision for future work be added.
Author Response
Comments 1: The introduction section should more prominently highlight the challenges and impacts of resource scarcity. This will better emphasize the necessity and urgency of the research, thus establishing a stronger rationale for the study.
Response 1: We added this in the introduction (6th and 7th paragraph): Despite vaccines being introduced many countries due to lack of resources were unable to combat the disease leading to massive breakouts13. Even with vaccines being approved by governments there is still a lack of diagnostic tools, therapeutics, and vaccines accessible to others. In fact, there is a low demand for vaccines worldwide due to accessibility as well as misinformation in the reported cases. COVID is still an issue as the number of cases being reported doesn’t represent the amount of people in the country who are suffering from COVID. The actual number of cases is between 2 to 20 times the amount being reported38. There have also been challenges in getting vaccines to countries that actually need them as many low-income countries have faced shortages in obtaining vaccines39. Overall, there was and still is a lack of resources due to less targeted responses and certain misconceptions leading to hesitancy towards vaccines
With fewer vaccines available in certain countries, new vaccine-resistant variants can emerge. The main purpose of vaccines is to limit the number of mutations a virus (in this case SARS-COV2) undergoes. When a country has fewer vaccines available, fewer people are vaccinated meaning there is a much greater chance that SARS-COV2 will have more mutations thus more variants. A lack of proper resources to combat the disease will lead to people getting more easily infected and not as quickly treated hence more people will suffer in those countries and the virus will spread quite rapidly.
If more emphasis is needed please let us know(that goes for any of the comments provided to us)
Comment 2: The analysis of existing research in the introduction is not sufficiently in-depth. There is a lack of systematic comparison between different methods and conclusions. It is recommended to include a detailed analysis of existing research, including a comparison of the advantages and disadvantages of various methods, and highlight the innovations in methodology or data application in this study to clarify its unique contributions.
Response 2: We included section 2 called related works: Previous works have focused significantly on predicting the spread of the COVID-19 virus 29. Hirschprung et al. focused on using multiple regression models and machine learning to predict the number of cases in countries around the world30. Wiezoreck et al. used a neural network-based deep learning architecture to achieve a forecasting accuracy of around 99%31. Works such as Tomar et al. focused more on specific countries such as India, and they were able to use a long short-term memory network to forecast the spread of the virus up to 90 days in advance 32. Finally, Yadav et al. also used statistical modeling methods to forecast the spread of COVID-19.
Fewer works have focused on individual countries’ risk predictions based on their vaccination data. Barda et al. focused on predicting the risk of death in certain countries in low-data situations 33. Similarly, Pal et al. used a neural network-based method to evaluate each country’s risk in the spread of COVID-19 34. Bird et al. also explored the effects of using various ensemble models on predicting the spread of the virus35. Finally, Chakraborty et al. developed a real-time forecast of COVID-19 cases to determine the risk of the virus. However, these works have not conducted a fully comparative analysis of different machine learning models on the most up-to-date data available.
Comment 3: The number of references is relatively low. It is suggested to increase the number of references from the relevant field to enhance the literature support and depth of the research background in the paper.
Response 3: We increased our references to 39. Would that be enough
Comment 4: The quality of the figures needs improvement. It is recommended to enhance the clarity and professionalism of the figures, especially Figure 3, which is not very clear.
Response 4: Are the names of the countries unclear? The image was imported directly from the tool we used to create the network (Gephi) so if you have any suggestions on how to make it more clear or professional please let us know.
Comment 5: The description of the dataset used in the paper lacks detail. It is recommended to provide more detailed information in this regard.
Response 5: We added a screenshot of our original and pre-processed dataset while also including the number of countries we collected data for as well as the number of vaccines we collected data for.
Comment 6: The analysis of the performance of the deep learning models used in the paper is not thorough enough, especially concerning accuracy and efficiency. It is suggested to strengthen the discussion in this area.
Response 6: We included the function we used to construct the neural network in the methodology. In the results and discussion area, we tried to use Pfizer as an example to help understand the network: To help understand how to interpret the networks we will use Pfizer as an example. Pfizer had 43 countries who used this vaccine and according to the 70-30 split, 30% of them equals 13. This means that the model tried to predict 13 countries. Using the accuracy rate (which we will discuss later), for Pfizer, it was around 97%, we can determine the amount the model predicted correctly and incorrectly (for Pfizer 12 were correct and 1 was incorrect). For the incorrect ones, it could be because there are countries close to the cut-off for the risk. The cutoff for Pfizer, for example, was at 50% so countries near 50% may have confused the model (there was one country that was around 47%). Another issue for other vaccines could be the amount of available data as some candidates only had 5 countries that used this vaccine thus there wasn’t enough data to train the model.
Comment 7: It was suggested that a vision for future work be added.
Response 7: We added another section at the end of our paper called future work: One improvement in this study could be to build the network on top of the world map allowing for a cleaner visualization while also maybe having the networks update live as well. Our networks and neural networks can also potentially be used to predict the countries that may struggle the most for a future disease based on stringency and vaccination percentages. These parameters can even be expanded and research can be done to determine other factors that may affect the disease. This can even be explored for COVID-19 to make this model more accurate. Examples of other factors could be the economic situation of certain countries or cultural biases toward vaccines.
Reviewer 2 Report
1. The writers have chosen the good work, whereas the models all exist. Already, many articles have been written using these models.
2. How is preprocessing done to compute the percentage of people vaccinated per country per manufacturer?
3. How to evaluate the risk of infection for each vaccine candidate.
1. Better to cite some more references in the introduction section.
2. The objectives of the work were not defined clearly. The dataset was also not mentioned clearly.
3. Need to elaborate on how metrics were evaluated.
4. In Section 2, the Methodology content was written in a General way. Need to define how the parameters were used in the model.
5. Explain clearly about Pfizer Network.
6. The results have to improve a lot. Reduce the content in the Conclusion section.
Author Response
Comment 1: Better to cite some more references in the introduction section.
Response: We added more citations and are up to 39
Comment 2: The objectives of the work were not defined clearly. The dataset was also not mentioned clearly.
Response: We tried to clarify the data by adding further descriptions and adding a screenshot of the original and pre-processed data in the methodology.
Comment 3: Need to elaborate on how metrics were evaluated.
Response: We expanded on how we pre-processed the data and the metrics we used to define the neural network
Comment 4: In Section 2, the Methodology content was written in a General way. Need to define how the parameters were used in the model.
Response 4: We included screenshots of our function as well as a few of the parameters in the methodology
Comment 5: Explain clearly about Pfizer Network.
Response 5: Under the image of our network we tried to explain Pfizer further
Comment 6: The results have to improve a lot. Reduce the content in the Conclusion section.
Response 6: We reduced the content in the conclusion and tried to improve the results by discussing Pfizer in more detail both in the context of the neural network and the network generated from Gephi
Round 2
Reviewer 1 Report
The author responded to my question seriously.
N/A
Author Response
Comment 1: The author responded to my question seriously.
Response 1: Thank you for all the suggestions!!
Reviewer 2 Report
1. The authors have done great work. However, they need to reduce the unnecessary content in the Introduction section and add some more content to the related work so that the authors can easily understand the existing problems and proposed work.
2. If possible, the proposed work implementation can be presented in the form of an algorithm.
3. In the results section, the visualization part requires high-resolution images.
Follow the Above suggestions
Author Response
Comment 1: The authors have done great work. However, they need to reduce the unnecessary content in the Introduction section and add some more content to the related work so that the authors can easily understand the existing problems and proposed work.
Response 1: We decreased the content in our Introduction but that did change the number of citations we had. We were however able to add on to the citations as we found 2-3 different research projects for the related work section. Hope our goal is more clear.
Comment 2: If possible, the proposed work implementation can be presented in the form of an algorithm.
Response 2: We made a flowchart for the methodology while adding a few more details for the network science section and changing the methodology format.
Comment 3: In the results section, the visualization part requires high-resolution images.
Response 3: We used a few websites to help our images have higher resolution so hope they are more clear
Thank you for all the comments
