AMP-GSM: Prediction of Antimicrobial Peptides via a Grouping–Scoring–Modeling Approach

Round 1

Reviewer 1 Report

Comments to the Author

 

This paper summarized the development of an in-silico-based model for the prediction of antimicrobial peptides. The Antimicrobial Peptide-Grouping-Scoring-Modeling (AMP-GSM) prediction model exhibited a good antimicrobial peptide classification performance and high prediction accuracy. In a word, this manuscript is innovative and practical. The construction process is also clearly expressed. However, several problems need to be addressed to meet the high standards of the Applied Sciences journal:

 

1.      The abstract section should provide the full version of the abbreviated phrases (e.g., AMP-GSM, AUC) to improve the readability of the texts.

2.      Line 15: Sentence structure: "computational" should be written in lowercase unless there is a particular reason as to why the first letter is in upper-case format

3.      Line 27-28: Revise the sentence structure for the AUC values obtained for both the gram-negative and gram-positive datasets. Instead of "%99" and "%98," it should be "99%" and "98%". Same for Line 239

4.      Line 29-30: There are many keywords, most of which are similar. For example, the keywords "grouping, scoring, modeling" are already covered by the term/word "machine learning." The same goes for "classification model" and "antimicrobial peptide prediction" Consider revising your keywords 5 or 6 keywords are adequate.

5.      Line 36-39: Please consider including references to studies that support the stated facts on the actions of AMPs against pathogens

6.      Consider revising Lines 56-61 for clarity. Chung et al. developed a prediction model based on RF. Then they compared its performance with other models built using different machine learning approaches using the same dataset for model training and performance analysis.

7.      Consider revising the graph in Figure 4. The y-axis values should be between 0 to 1 (0 to 100%), yet there is a "1.2" label in the graph

8.      In my opinion, Figure 5 needs to be revised, although earlier results in the manuscript reported the robust performance of the AMP-GSM prediction model compared to traditional QSAR models. From the earlier results in Table 10. The AMP-GSM model had Sn of 91.01(±0.23), Sp 92.97(±0.03), balanced accuracy of 91.71 (±0.13), and auROC of 97.07(±0.06), yet in Figure 5 it seems like it has 100% values for all those four parameters. With any model, though, one is never going to hit 100% accuracy.

 

General remark in terms of grammar: The manuscript needs a minor revision to improve readability:

 

-          For example, in Lines 44-45, it's advisable to consider removing "etc." at the end of the sentence and putting an "and" after LAMP [5]. The same case for Lines 53-54

-          Line 61-62: "In this model, Using the AMP datasets obtained from different...." placement of uppercase letters in the middle of the sentence

-          Line 58: "k-NN algorithms," and the full version of the abbreviated phrase is in Line 60. Please consider mentioning the full-word version in line 58, where the kNN is first mentioned

-          Grammar Line 100: "score by using the features in which group or groups in antimicrobial peptide prediction." Should it be "in each group or groups...."?

-          Title for Table 1: The first letter should be in uppercase

-          Consider revising the Table captions for all the Tables in the manuscript. There are considerably too detailed. Table captions should be short. In addition, you can put a footnote at the bottom of the table to

-          Grammar correction "Performans" should be "Performance" for Tables 2, 4, 5, and 6.

-          Table 6: The headings terms can be abbreviated to display the table layout better. For example, Specificity is Sp; sensitivity is Sn; Area Under Curve is AUC. The full version can be placed in the footnote

-          Table 7 insert a footnote at the bottom to define the "FS method", "ML Method", "AUC", "F1"

-          In some parts of the manuscript, the first letter of the words "gram-negative and gram-positive" is in uppercase. E.g., Line 378-379; Line 382, while in other parts of the manuscript, the same first letters are lowercase, e.g., Line 385. Is there any particular reason for this? Consider revising.

 

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 2 Report

The authors have selected an important and relevant topic as the emergence of antimicrobial resistance is an ever growing challenge to the medical fraternity.

The manuscript is well written and can be accepted for publication after minor modifcations:

1. The abstract is too concise, it must be elaborated to focus the important findings of this study and highlight its future implications.

2. The are spelling and grammatical errors that need to be addressed.

3. Discussion: This part is weak and needs to be reworked; the majority of the content is just introductory, and there is little comparison of the data obtained to previous work, reference materials, or similar examples (previously published similar articles) must be cited.

 

4.       Include the limitations, and future implications of this study before the “Conclusion” part.

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx