Dear Mr Bouhadjar,

Thank you very much for submitting your manuscript "Sequence learning, prediction, and replay in networks of spiking neurons" for consideration at PLOS Computational Biology.

As with all papers reviewed by the journal, your manuscript was reviewed by members of the editorial board and by several independent reviewers. In light of the reviews (below this email), we would like to invite the resubmission of a significantly-revised version that takes into account the reviewers' comments.

We cannot make any decision about publication until we have seen the revised manuscript and your response to the reviewers' comments. Your revised manuscript is also likely to be sent to reviewers for further evaluation.

When you are ready to resubmit, please upload the following:

[1] A letter containing a detailed list of your responses to the review comments and a description of the changes you have made in the manuscript. Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

[2] Two versions of the revised manuscript: one with either highlights or tracked changes denoting where the text has been changed; the other a clean version (uploaded as the manuscript file).

Important additional instructions are given below your reviewer comments.

Please prepare and submit your revised manuscript within 60 days. If you anticipate any delay, please let us know the expected resubmission date by replying to this email. Please note that revised manuscripts received after the 60-day due date may require evaluation and peer review similar to newly submitted manuscripts.

Thank you again for your submission. We hope that our editorial process has been constructive so far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

Stefan Kiebel

Associate Editor

PLOS Computational Biology

Samuel Gershman

Deputy Editor

PLOS Computational Biology

***********************

Reviewer's Responses to Questions

Comments to the Authors:

Reviewer #1: The review is uploaded as an attachment

Reviewer #2: The paper from Bouhadjar et al. is asking an interesting question. The authors wonder how to combine several ‘universal’ tasks in the same network in a more biological implementation. In previous studies, these tasks were either studied separately, or in a somewhat artificial set-up. In their framework the homeostasis in particular, enhancing context specificity, has great merit. I would like to ask the authors, before any final recommendation is made, to answer some questions I had when reading the manuscript. In general the manuscript is of good quality. I also am happy that the authors are very honest about the limitations of the model, such as the limited replay capability.

Some major questions:

While combining several tasks is a feature of the model, previous work focusing on individual tasks may have the benefit that they can zoom in and mechanistically understand things in great detail. It seems that some part of the solution or implementation in these other studies could be shared with the solution in this paper. In this regard, you already cited resemblance with amongst others Cone and Shouval, eLife 2021, the sequence learning aspect. A more detailed discussion may be beneficial for the reader. As the authors are also interested in mismatch/prediction it would be good to discuss resemblance in this domain, for example: Schulz et al., eLife 2021. Finally I am wondering if there is any link to feature segmentation, as studied by for example the group of Tomoki Fukai (example paper is Asabuki and Fukai, Nat Comms 2020).

Is there a limit to the history/context memory? For example would ABCDEFG and HBCDEFJ give different predictions for the final element? In other words, would the homeostasis still be able to separate the pathways?

Some minor questions:

Network structure: the M non-overlapping subpopulations are hardcoded into the exc neurons rnn? This merely means that they receive the same input, but the connectivity is still random?

External inputs: how long in time are individual elements? When for example sequence ABCD is given, does each letter trigger a single spike? What if it triggers multiple? I suppose the refractory period does not allow this.

Equation 1: x_j is written after the summation and y_i before. For esthetic reasons I would put both either before or after.

Line 223: why are the minimal permanences uniformly distributed? Is this how they are initialized, and hence guarantee the necessary heterogeneity?

Are the parameters, for example dAP threshold of 59pA etc, taken from a particular previous study or finetuned?

What is the purpose of the recurrent EE connectivity? Am I correct in saying that the feedforward pathways, being carved out, is what makes the model work?

**********

Have the authors made all data and (if applicable) computational code underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data and code underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data and code should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data or code —e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

Figure Files:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

Data Requirements:

Please note that, as a condition of publication, PLOS' data policy requires that you make available all data used to draw the conclusions outlined in your manuscript. Data must be deposited in an appropriate repository, included within the body of the manuscript, or uploaded as supporting information. This includes all numerical values that were used to generate graphs, histograms etc.. For an example in PLOS Biology see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5.

Reproducibility:

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Attachments

Attachment

Submitted filename: PCOMPBIOLD210201_review.pdf

Dear Mr Bouhadjar,

Thank you very much for submitting your manuscript "Sequence learning, prediction, and replay in networks of spiking neurons" for consideration at PLOS Computational Biology.

In light of the reviews (below this email), we would like to invite the resubmission of a significantly-revised version that takes into account the reviewers' comments.

As you can see, the remaining reviewer made the point that there is a gap between your model features and key features of experimentally observed neuronal activity. I concur with the reviewer that although your efforts to implement the HTM in a spiking system are laudable, from the current results  it not clear how helpful the model may be in the future for experimental studies. This is important because the scope of PLOS Computational Biology is aimed towards biological discovery through modelling. Comments in this direction were made by the reviewer in the first rounds of review and as far as I can see you added text to the discussion but didn't consider the reviewer's suggestions towards adding simulations. Therefore, I strongly recommend that you address these comments about suggested further simulations, e.g. to test the model with some background activity and rate modulation of the input.

We cannot make any decision about publication until we have seen the revised manuscript and your response to the reviewers' comments. Your revised manuscript is also likely to be sent to reviewers for further evaluation.

When you are ready to resubmit, please upload the following:

[1] A letter containing a detailed list of your responses to the review comments and a description of the changes you have made in the manuscript. Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out.

[2] Two versions of the revised manuscript: one with either highlights or tracked changes denoting where the text has been changed; the other a clean version (uploaded as the manuscript file).

Important additional instructions are given below your reviewer comments.

Please prepare and submit your revised manuscript within 60 days. If you anticipate any delay, please let us know the expected resubmission date by replying to this email. Please note that revised manuscripts received after the 60-day due date may require evaluation and peer review similar to newly submitted manuscripts.

Thank you again for your submission. We hope that our editorial process has been constructive so far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

Stefan Kiebel

Associate Editor

PLOS Computational Biology

Samuel Gershman

Deputy Editor

PLOS Computational Biology

***********************

Reviewer's Responses to Questions

Comments to the Authors:

Reviewer #1: The authors propose a new implementation of the Hierarchical Temporal Memory (HTM) algorithm in terms of biologically interpretable neural networks. The model has some interesting and valuable ideas but some of the simplifications applied in the simulations and model limit in my eyes its impact.

As outlined in the major points of my previous review, the activity produced by the model is extremely far from what is observed experimentally. Input stimuli are passed to the network by one single spike; there is no rate modulation, neither in the input signal nor in the response of the network; and the model does not display any background activity. These are extreme conditions, and leave the reader wondering how the model would function in a more realistic scenario. Therefore, I had suggested new simulations.

For example, a first quick way to address some of these concerns could have been to re-run the existing model (learning and prediction) with input signals that include (beyond the sequences to learn) also noise spikes and/or rate modulation. Such analyses are quick to perform, as they do not require any modification in the model, and their outcome, either positive or negative, would be highly informative.

The authors did not perform new analyses but addressed the concerns by adding a paragraph in the Discussion of the manuscript (and a new Suppl. Fig. 5) where they discussed the degree of synchronization present in experimental data and how synchrony might go unnoticed when subsampling.

In their reply letter the Authors write: “In a new paragraph added to “Discussion: Limitations and outlook”, we point at the apparent mismatch between the firing statistics in our model and those found in nature, and explain how a more natural irregular activation of sequences, parallel processing, and subsampling would resolve this mismatch”. In my opinion, this does not resolve the mismatch and does not address the concern raised. Moreover, this line of argumentation (together with the addition of Suppl. Fig. 5) appears to suggest that brain activity 'is' exclusively composed of synfire chains. While there is large experimental evidence of the presence of highly synchronized neuronal ensemble in the brain, assuming that such kind of activity constitutes the totality of the neuronal activity is a very strong claim.

Therefore, in my opinion it is necessary to perform additional simulations, on the line of those described above, to provide insightful addition to the field suited for PLOS Computational Biology.

**********

Have the authors made all data and (if applicable) computational code underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data and code underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data and code should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data or code —e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Figure Files:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

Data Requirements:

Please note that, as a condition of publication, PLOS' data policy requires that you make available all data used to draw the conclusions outlined in your manuscript. Data must be deposited in an appropriate repository, included within the body of the manuscript, or uploaded as supporting information. This includes all numerical values that were used to generate graphs, histograms etc.. For an example in PLOS Biology see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5.

Reproducibility:

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

Dear Mr Bouhadjar,

Thank you very much for submitting your manuscript "Sequence learning, prediction, and replay in networks of spiking neurons" for consideration at PLOS Computational Biology. As with all papers reviewed by the journal, your manuscript was reviewed by members of the editorial board and by several independent reviewers. The reviewers appreciated the attention to an important topic. Based on the reviews, we are likely to accept this manuscript for publication, providing that you modify the manuscript according to the review recommendations.

Please prepare and submit your revised manuscript within 30 days. If you anticipate any delay, please let us know the expected resubmission date by replying to this email.

When you are ready to resubmit, please upload the following:

[1] A letter containing a detailed list of your responses to all review comments, and a description of the changes you have made in the manuscript. Please note while forming your response, if your article is accepted, you may have the opportunity to make the peer review history publicly available. The record will include editor decision letters (with reviews) and your responses to reviewer comments. If eligible, we will contact you to opt in or out

[2] Two versions of the revised manuscript: one with either highlights or tracked changes denoting where the text has been changed; the other a clean version (uploaded as the manuscript file).

Important additional instructions are given below your reviewer comments.

Thank you again for your submission to our journal. We hope that our editorial process has been constructive so far, and we welcome your feedback at any time. Please don't hesitate to contact us if you have any questions or comments.

Sincerely,

Stefan Kiebel

Associate Editor

PLOS Computational Biology

Samuel Gershman

Deputy Editor

PLOS Computational Biology

***********************

A link appears below if there are any accompanying review attachments. If you believe any reviews to be missing, please contact ploscompbiol@plos.org immediately:

[LINK]

Reviewer's Responses to Questions

Comments to the Authors:

Please note here if the review is uploaded as an attachment.

Reviewer #1: I appreciate the new simulations performed by the Authors, which can better inform the readers about the response range of the model.

I would suggest the publication of the current paper pending one final change.

On lines 639-640 the authors state: ‘As shown in Fig. S6 and Fig. S7, the spiking activity and the prediction performance of the TM model are robust with respect to low and moderate levels of synaptic background activity'.

In figure S6 are shown results for a no-noise condition (panel A) and for two different degrees of noise (panels B and C). In figure S7 it is shown that for the second of these two noise conditions the network loses its ability to learn sequences.

Given these results, I would kindly ask the authors to remove from the previous sentence the words “and moderate”.

Firstly, only two levels of noise have been tested and for only one of these two the network maintained the desired qualities. Hence, it is not clear to what refers the sentence ‘robust with respect to low and moderate levels of synaptic background activity’. Secondly, the model works only until 'non-task related background spikes' are generated (as confirmed by the authors), which is over all a clearly very low noise condition (as also visible in Fig. S6 B).

**********

Have the authors made all data and (if applicable) computational code underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data and code underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data and code should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data or code —e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

**********

PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Figure Files:

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email us at figures@plos.org.

Data Requirements:

Please note that, as a condition of publication, PLOS' data policy requires that you make available all data used to draw the conclusions outlined in your manuscript. Data must be deposited in an appropriate repository, included within the body of the manuscript, or uploaded as supporting information. This includes all numerical values that were used to generate graphs, histograms etc.. For an example in PLOS Biology see here: http://www.plosbiology.org/article/info%3Adoi%2F10.1371%2Fjournal.pbio.1001908#s5.

Reproducibility:

To enhance the reproducibility of your results, we recommend that you deposit your laboratory protocols in protocols.io, where a protocol can be assigned its own identifier (DOI) such that it can be cited independently in the future. Additionally, PLOS ONE offers an option to publish peer-reviewed clinical study protocols. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols

References:

Review your reference list to ensure that it is complete and correct. If you have cited papers that have been retracted, please include the rationale for doing so in the manuscript text, or remove these references and replace them with relevant current references. Any changes to the reference list should be mentioned in the rebuttal letter that accompanies your revised manuscript.

If you need to cite a retracted article, indicate the article’s retracted status in the References list and also include a citation and full reference for the retraction notice.

Dear Mr Bouhadjar,

We are pleased to inform you that your manuscript 'Sequence learning, prediction, and replay in networks of spiking neurons' has been provisionally accepted for publication in PLOS Computational Biology.

Before your manuscript can be formally accepted you will need to complete some formatting changes, which you will receive in a follow up email. A member of our team will be in touch with a set of requests.

Please note that your manuscript will not be scheduled for publication until you have made the required changes, so a swift response is appreciated.

IMPORTANT: The editorial review process is now complete. PLOS will only permit corrections to spelling, formatting or significant scientific errors from this point onwards. Requests for major changes, or any which affect the scientific understanding of your work, will cause delays to the publication date of your manuscript.

Should you, your institution's press office or the journal office choose to press release your paper, you will automatically be opted out of early publication. We ask that you notify us now if you or your institution is planning to press release the article. All press must be co-ordinated with PLOS.

Thank you again for supporting Open Access publishing; we are looking forward to publishing your work in PLOS Computational Biology. 

Best regards,

Stefan Kiebel

Associate Editor

PLOS Computational Biology

Samuel Gershman

Deputy Editor

PLOS Computational Biology

***********************************************************