Comparing the Pfizer Central Nervous System Multiparameter Optimization Calculator and a BBB Machine Learning Model
The ability to calculate whether small molecules will cross the blood-brain barrier (BBB) is an important task for companies working in neuroscience drug discovery. For a decade, scientists have relied on relatively simplistic rules such as Pfizer's central nervous system multiparameter optimiz...
Saved in:
| Published in: | ACS chemical neuroscience Vol. 12; no. 12; p. 2247 |
|---|---|
| Main Authors: | , , , |
| Format: | Journal Article |
| Language: | English |
| Published: |
16.06.2021
|
| ISSN: | 1948-7193, 1948-7193 |
| Online Access: | Get more information |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Abstract | The ability to calculate whether small molecules will cross the blood-brain barrier (BBB) is an important task for companies working in neuroscience drug discovery. For a decade, scientists have relied on relatively simplistic rules such as Pfizer's central nervous system multiparameter optimization models (CNS-MPO) for guidance during the drug selection process. In parallel, there has been a continued development of more sophisticated machine learning models that utilize different molecular descriptors and algorithms; however, these models represent a "black box" and are generally less interpretable. In both cases, these methods predict the ability of small molecules to cross the BBB using the molecular structure information on its own without in vitro or in vivo data. We describe here the implementation of two versions of Pfizer's algorithm (Pf-MPO.v1 and Pf-MPO.v2) and compare it with a Bayesian machine learning model of BBB penetration trained on a data set of 2296 active and inactive compounds using extended connectivity fingerprint descriptors. The predictive ability of these approaches was compared with 40 known CNS active drugs initially used by Pfizer as their positive set for validation of the Pf-MPO.v1 score. 37/40 (92.5%) compounds were predicted as active by the Bayesian model, while only 30/40 (75%) received a desirable Pf-MPO.v1 score ≥4 and 33/40 (82.5%) received a desirable Pf-MPO.v2 score ≥4, suggesting the Bayesian model is more accurate than MPO algorithms. This also indicates machine learning models are more flexible and have better predictive power for BBB penetration than simple rule sets that require multiple, accurate descriptor calculations. Our machine learning model statistics are comparable to recent published studies. We describe the implications of these findings and how machine learning may have a role alongside more interpretable methods.The ability to calculate whether small molecules will cross the blood-brain barrier (BBB) is an important task for companies working in neuroscience drug discovery. For a decade, scientists have relied on relatively simplistic rules such as Pfizer's central nervous system multiparameter optimization models (CNS-MPO) for guidance during the drug selection process. In parallel, there has been a continued development of more sophisticated machine learning models that utilize different molecular descriptors and algorithms; however, these models represent a "black box" and are generally less interpretable. In both cases, these methods predict the ability of small molecules to cross the BBB using the molecular structure information on its own without in vitro or in vivo data. We describe here the implementation of two versions of Pfizer's algorithm (Pf-MPO.v1 and Pf-MPO.v2) and compare it with a Bayesian machine learning model of BBB penetration trained on a data set of 2296 active and inactive compounds using extended connectivity fingerprint descriptors. The predictive ability of these approaches was compared with 40 known CNS active drugs initially used by Pfizer as their positive set for validation of the Pf-MPO.v1 score. 37/40 (92.5%) compounds were predicted as active by the Bayesian model, while only 30/40 (75%) received a desirable Pf-MPO.v1 score ≥4 and 33/40 (82.5%) received a desirable Pf-MPO.v2 score ≥4, suggesting the Bayesian model is more accurate than MPO algorithms. This also indicates machine learning models are more flexible and have better predictive power for BBB penetration than simple rule sets that require multiple, accurate descriptor calculations. Our machine learning model statistics are comparable to recent published studies. We describe the implications of these findings and how machine learning may have a role alongside more interpretable methods. |
|---|---|
| AbstractList | The ability to calculate whether small molecules will cross the blood-brain barrier (BBB) is an important task for companies working in neuroscience drug discovery. For a decade, scientists have relied on relatively simplistic rules such as Pfizer's central nervous system multiparameter optimization models (CNS-MPO) for guidance during the drug selection process. In parallel, there has been a continued development of more sophisticated machine learning models that utilize different molecular descriptors and algorithms; however, these models represent a "black box" and are generally less interpretable. In both cases, these methods predict the ability of small molecules to cross the BBB using the molecular structure information on its own without in vitro or in vivo data. We describe here the implementation of two versions of Pfizer's algorithm (Pf-MPO.v1 and Pf-MPO.v2) and compare it with a Bayesian machine learning model of BBB penetration trained on a data set of 2296 active and inactive compounds using extended connectivity fingerprint descriptors. The predictive ability of these approaches was compared with 40 known CNS active drugs initially used by Pfizer as their positive set for validation of the Pf-MPO.v1 score. 37/40 (92.5%) compounds were predicted as active by the Bayesian model, while only 30/40 (75%) received a desirable Pf-MPO.v1 score ≥4 and 33/40 (82.5%) received a desirable Pf-MPO.v2 score ≥4, suggesting the Bayesian model is more accurate than MPO algorithms. This also indicates machine learning models are more flexible and have better predictive power for BBB penetration than simple rule sets that require multiple, accurate descriptor calculations. Our machine learning model statistics are comparable to recent published studies. We describe the implications of these findings and how machine learning may have a role alongside more interpretable methods.The ability to calculate whether small molecules will cross the blood-brain barrier (BBB) is an important task for companies working in neuroscience drug discovery. For a decade, scientists have relied on relatively simplistic rules such as Pfizer's central nervous system multiparameter optimization models (CNS-MPO) for guidance during the drug selection process. In parallel, there has been a continued development of more sophisticated machine learning models that utilize different molecular descriptors and algorithms; however, these models represent a "black box" and are generally less interpretable. In both cases, these methods predict the ability of small molecules to cross the BBB using the molecular structure information on its own without in vitro or in vivo data. We describe here the implementation of two versions of Pfizer's algorithm (Pf-MPO.v1 and Pf-MPO.v2) and compare it with a Bayesian machine learning model of BBB penetration trained on a data set of 2296 active and inactive compounds using extended connectivity fingerprint descriptors. The predictive ability of these approaches was compared with 40 known CNS active drugs initially used by Pfizer as their positive set for validation of the Pf-MPO.v1 score. 37/40 (92.5%) compounds were predicted as active by the Bayesian model, while only 30/40 (75%) received a desirable Pf-MPO.v1 score ≥4 and 33/40 (82.5%) received a desirable Pf-MPO.v2 score ≥4, suggesting the Bayesian model is more accurate than MPO algorithms. This also indicates machine learning models are more flexible and have better predictive power for BBB penetration than simple rule sets that require multiple, accurate descriptor calculations. Our machine learning model statistics are comparable to recent published studies. We describe the implications of these findings and how machine learning may have a role alongside more interpretable methods. |
| Author | Zorn, Kimberley M Brunner, Daniela Urbina, Fabio Ekins, Sean |
| Author_xml | – sequence: 1 givenname: Fabio surname: Urbina fullname: Urbina, Fabio – sequence: 2 givenname: Kimberley M surname: Zorn fullname: Zorn, Kimberley M – sequence: 3 givenname: Daniela surname: Brunner fullname: Brunner, Daniela – sequence: 4 givenname: Sean surname: Ekins fullname: Ekins, Sean |
| BookMark | eNpNjz1PwzAYhC0EEm3hHzB4ZEmxncSJRxrxJbUUCZirt84bauTYxXaQ6K8nCAamO50e3emm5Nh5h4RccDbnTPAr0FHvsHc4BD_nmjEhyyMy4aqos4qr_PifPyXTGN8Zk4rVckJi4_s9BOPeaNohferMAQNt0KUAlj5i-PRDpM9fMWFPV4NNZqShxzRS630yvTlAMt7RBqweLCQfKLiWAl0sFnQFemcc0iVCcD8bK9-iPSMnHdiI5386I6-3Ny_NfbZc3z0018sMCiVSJmvR6q4sUJajKbTaCi3bTucSFORtJWqpsEMuKybHcDsirNSlapXSWlZczMjlb-8--I8BY9r0Jmq0FhyOrzaizHmZ55wL8Q0HTGVe |
| CitedBy_id | crossref_primary_10_1021_acsptsci_5c00299 crossref_primary_10_1080_17460441_2023_2294118 crossref_primary_10_1016_j_talanta_2022_123396 crossref_primary_10_3390_ijms26167976 crossref_primary_10_3389_fcimb_2022_838259 crossref_primary_10_3390_ijms25179386 crossref_primary_10_1021_acs_jcim_5c00590 crossref_primary_10_1016_j_ailsci_2025_100143 crossref_primary_10_1016_j_neurot_2025_e00624 |
| ContentType | Journal Article |
| DBID | 7X8 |
| DOI | 10.1021/acschemneuro.1c00265 |
| DatabaseName | MEDLINE - Academic |
| DatabaseTitle | MEDLINE - Academic |
| DatabaseTitleList | MEDLINE - Academic |
| Database_xml | – sequence: 1 dbid: 7X8 name: MEDLINE - Academic url: https://search.proquest.com/medline sourceTypes: Aggregation Database |
| DeliveryMethod | no_fulltext_linktorsrc |
| Discipline | Anatomy & Physiology |
| EISSN | 1948-7193 |
| GroupedDBID | --- 53G 55A 5VS 6J9 7X8 7~N AABXI AAKDD ABBLG ABJNI ABLBI ABMVS ABQRX ABUCX ACGFO ACGFS ACS ADBBV ADHLV AEESW AENEX AFEFF AHGAQ ALMA_UNASSIGNED_HOLDINGS AQSVZ BAANH BAWUL CUPRZ DIK EBS ED~ F5P GGK GNL GX1 IH9 JG~ OK1 RNS ROL RPM UI2 VF5 VG9 W1F |
| ID | FETCH-LOGICAL-a492t-682dcf54e652dc4c9b2c6dfc36a9a3d72869efe16706c36bc4c05c59d99cc6712 |
| IEDL.DBID | 7X8 |
| ISICitedReferencesCount | 20 |
| ISICitedReferencesURI | http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000664290200017&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| ISSN | 1948-7193 |
| IngestDate | Fri Jul 11 11:52:39 EDT 2025 |
| IsDoiOpenAccess | false |
| IsOpenAccess | true |
| IsPeerReviewed | true |
| IsScholarly | true |
| Issue | 12 |
| Language | English |
| LinkModel | DirectLink |
| MergedId | FETCHMERGED-LOGICAL-a492t-682dcf54e652dc4c9b2c6dfc36a9a3d72869efe16706c36bc4c05c59d99cc6712 |
| Notes | ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23 |
| OpenAccessLink | https://www.ncbi.nlm.nih.gov/pmc/articles/8260158 |
| PQID | 2531533112 |
| PQPubID | 23479 |
| ParticipantIDs | proquest_miscellaneous_2531533112 |
| PublicationCentury | 2000 |
| PublicationDate | 2021-06-16 |
| PublicationDateYYYYMMDD | 2021-06-16 |
| PublicationDate_xml | – month: 06 year: 2021 text: 2021-06-16 day: 16 |
| PublicationDecade | 2020 |
| PublicationTitle | ACS chemical neuroscience |
| PublicationYear | 2021 |
| SSID | ssj0069086 |
| Score | 2.3585258 |
| Snippet | The ability to calculate whether small molecules will cross the blood-brain barrier (BBB) is an important task for companies working in neuroscience drug... |
| SourceID | proquest |
| SourceType | Aggregation Database |
| StartPage | 2247 |
| Title | Comparing the Pfizer Central Nervous System Multiparameter Optimization Calculator and a BBB Machine Learning Model |
| URI | https://www.proquest.com/docview/2531533112 |
| Volume | 12 |
| WOSCitedRecordID | wos000664290200017&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D |
| hasFullText | |
| inHoldings | 1 |
| isFullTextHit | |
| isPrint | |
| link | http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV09T8MwELWAMrDwVRDfOiTEFpo4sR1PqK2oWCgdQOpWOY5TVWpSaAoS_HrOrgsDCxJbFHlIrPP53d27e4RcUakyxUSGyE3abJW2fjDMAqp1kuQil6kOndiE6PfT4VAOfMKt9rTKlU90jjqfaZsjb1E0FoQmCA9uX14Dqxplq6teQmOdNGKEMpbSJYbfVQQM_JzSI8bpaSAQqaxa52jUUhqDR1O6oZE3kbahCPvljt0d09v579ftkm2PLqG9NIc9smaqfdJsVxhZlx9wDY7v6RLpTVJ3lxKE1RgQBMKgmHyaOfhsL_TRh8zealhONAfXp2vHhJeWPgOP6GhK38EJXTW1ZFYM3kFVOSjodDrw4EiaBvz81jFY0bXpAXnu3T117wMvwRCoRNJFwFOa64IlhjN8SLTMqOZ5oWOupIpzQVMuTWEiLkKOLzNcEjLNZC6l1lxE9JBsVLPKHBEIBVVxyrThmUhSvAUR-UQFCxWeU5UwekwuV1s7QhO3dQtVGfzT0c_mnvxhzSnZopZ3YvWF-BlpFHiMzTnZ1O-LST2_cBbyBaldx2g |
| linkProvider | ProQuest |
| openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Comparing+the+Pfizer+Central+Nervous+System+Multiparameter+Optimization+Calculator+and+a+BBB+Machine+Learning+Model&rft.jtitle=ACS+chemical+neuroscience&rft.au=Urbina%2C+Fabio&rft.au=Zorn%2C+Kimberley+M&rft.au=Brunner%2C+Daniela&rft.au=Ekins%2C+Sean&rft.date=2021-06-16&rft.issn=1948-7193&rft.eissn=1948-7193&rft.volume=12&rft.issue=12&rft.spage=2247&rft_id=info:doi/10.1021%2Facschemneuro.1c00265&rft.externalDBID=NO_FULL_TEXT |
| thumbnail_l | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1948-7193&client=summon |
| thumbnail_m | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1948-7193&client=summon |
| thumbnail_s | http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1948-7193&client=summon |