Patent Information Users Group, Inc.

American Chemical Society Spring 2021 National Meeting: selected presentations of  [CHAL] Division of Chemistry & the Law (Apr.8-9, 2021) and [CINF] Division of Chemical Information, (Apr. 12-16, 2021) symposia

Because of size limitations, the post is split in two parts: Part I [CHAL] Division of Chemistry & the Law (Apr.8-9, 2021) & [CINF] Division of Chemical Information, (Apr. 12-13, 2021); Part II [CINF] Division of Chemical Information, (Apr. 14-16, 2021)

Technical Program website: https://acs.digitellinc.com/acs/live/8/page/18

To read abstracts, follow links (from a title) and choose a presentation by time (Change from PT to ET)  - LInkes to live program are not valid after time of presentations.

Full CINF technical program have been published in Chemical Information Bulletin 2021, 73  (1), pp. 30-47 (Pacific Time)

Note: Short annotations and authors affiliations would be added later which may clarify importance/interest of selected presentations.

Registration page (Regular registration fee $99 for ACS Members and $149 for non-members)

Division/Committee: [CHAL] Division of Chemistry & the Law

April 8, 2021 Thursday

Tips for requesting chemical searches
04:00pm - 04:30pm USA / Canada - Eastern - April 8, 2021
Elaine Cheeseman (CAS), Presenter
Track: [CHAL] Division of Chemistry & the Law
This presentation will provide guidance for chemical searches that will help the client and searcher collaboratively align the requested project objectives.

Updates on recent federal circuit court decisions
05:00pm - 05:30pm USA / Canada - Eastern - April 8, 2021
Andrew Berks, Ph.D., J.D.,(Gallet Dreyer & Berkey, LLP), Presenter
Track: [CHAL] Division of Chemistry & the Law
https://acs.digitellinc.com/acs/live/8/page/18/1?eventSearchInput=&eventSearchDate=2021-04-08&eventSearchTrack=63&eventSearchTag=0

April 9, 2021, Friday

The Hatch Waxman Act: An overview and intellectual property considerations thereof
02:00pm - 02:30pm USA / Canada - Eastern - April 9, 2021
Sameshnee Pelly, Presenter (Finnegan Henderson Farabow Garrett and Dunner LLP)
Track: [CHAL] Division of Chemistry & the Law
This presentation will provide a brief overview of the Hatch-Waxman Act [“Drug Price Competition and Patent Term Restoration Act of 1984,”] and intellectual property considerations such as patent listing requirements, paragraph IV certifications, and pharmaceutical patent litigation.
See  also :
Schacht, W.H., Thomas, J.R., 2012. The Hatch-Waxman Act: A Quarter Century Later (CRS Report No. R41114), Congressional Research Service. 21 p.(Dec. 05, 2012 Revision) https://www.everycrsreport.com/reports/R41114.html

06:05pm - 06:30pm USA / Canada - Eastern - April 9, 2021
John Landells, Andrew Gregory & Declan McKeveney, Presenters   [FB Rice Pty Ltd, Australia]
Track: [CHAL] Division of Chemistry & the Law
…This talk focuses initially on the generous grace period provisions available in Australia. A comparison is then made with the relatively new grace period allowances in New Zealand and those available in the increasingly important Southeast Asia region.
See also:
Gumley, T., Mok, J.( FPA Patent Attorners) 2014. Patent grace periods in South East Asia. URL https://fpapatents.com/resource?id=309 (accessed 4.3.21).
Nugent, Brendan (Michael Buck IP), 2019. Grace Periods in Australia and New Zealand. URL https://www.patentsandtrademarks.com.au/grace-periods-in-australia-and-new-zealand/ (accessed 4.3.21).
Quinn, Edward, November 2020. Extension of time for grace period in Australia. Spruson & Ferguson. URL https://www.spruson.com/patents/extension-of-time-for-grace-period-in-australia/ (accessed 4.3.21).

Improving patent analytics using machine learning technologies and introducing ML4Patents.com

Anthony Trippe, Presenter [Patinformatics LLC, ML4Patents.com]
…Starting with a general overview of patent analytics, and with a focus on patent landscape reports, case studies and perspectives will be provided on why this work is so highly valued. The presentation will conclude with a prioritized list of suggestions for how patent analytics and patent landscape creation could be aided by the further development and implementation of semantic technologies.
See also
ML4Patents.com (the industry's online resource for the use of machine learning and artificial intelligence in the field of Intellectual Property)

[CINF] Division of Chemical Information

April 12, 2021,  Monday

Symposium:
Innovations in Open Data Exchange and Information Dissemination:  
12:00pm - 03:00pm USA / Canada - Eastern - April 12, 2021
Sunghwan Kim, Organizer, Presider  [NIH]  |     Wendy Patterson, Organizer, Presider    |     Susan Cardinal, Organizer, Presider
Track: [CINF] Division of Chemical Information

Pharos: An open data resource for examining target, disease and ligand interactions
02:15pm - 02:35pm USA / Canada - Eastern - April 12, 2021
Timothy Sheils, Presenter    |     Keith Kelleher    |     Dac-Trung Nguyen [National Center for Advancing Translational Sciences/NIH, Frederick, MD]
Track: [CINF] Division of Chemical Information
Pharos is a web resource that originated from the NIH "Illuminating the Druggable Genome" (IDG) program to characterize the understudied [so-called “dark”] regions of the druggable genome... Pharos is built on top of the Target Central Resource Database (TCRD), and the most recent update for TCRD now incorporates 78 disparate data sources, incorporating target expression, protein-protein interaction, disease ontology and chemical activity databases. … Pharos now enables users to explore subsets of related targets, as well as view their similarity overlap. … By continually incorporating new data sets, Pharos allows for more unique combinations of data and facilitates the discovery of interesting related targets and shows connections from diseases to targets to ligands.
See also:
Pharos (https://pharos.nih.gov/) Targets (20412),  Knowledge Metrics: PPIs(Protein-protein interactions). PubMed Score Jensen Lab score, prevalence of the target in PubMed Articles; PubTator Score (based on PubMed textmining), Antibody Count [for a target, in antibodypedia], Log Novelty [Tin-X metrics on scarcity publication on the target]; Diseases (19683); Ligands (266944)
Sheils, T.K., Mathias, S.L., Kelleher, K.Jet. al. 2021. TCRD and Pharos 2021: mining the human proteome for disease biology. Nucleic Acids Research 49 (D1), D1334–D1346. https://doi.org/10.1093/nar/gkaa993

 

Synergy through integration of data sources
02:35pm - 02:55pm USA / Canada - Eastern - April 12, 2021
Ian Wetherbee, Presenter [Google]   |     Stephen Boyer, Presenter  [Collabra Inc, San Jose, CA]  |     Lutz Weber  [OntoChem, Germany]  |     Bob Frommer  [Collabra Inc]  |     Jane Frommer [Collabra Inc]  
Track: [CINF] Division of Chemical Information
[For the power user]...[Without need building and maintaining in-house data repositories] [w]ith a common storage and analysis platform with the right cost-sharing model, the scientific community can collaboratively expand and access seemingly unrelated datasets and pose intradisciplinary questions relevant to the advancement of science. Examples:…identifying molecules in litigation, associating government agencies with specific drug- and disease-funding, tracking clinical trials with chemical structures in patents, grant applications and in scientific documents. In semantic aggregation, tagging compounds according to their structure and properties allows association of knowledge from different data sources to be cross-correlated as well as fed into machine learning and prediction systems.
See also:
Patent analysis using the Google Patents Public Datasets on BigQuery[github.com],
Google Patents Public Datasets [console.cloud.google.com]  (24 free/paid public datasets: Google Patents Public Data (IFI CLAIMS), Google Patents Research Data (GooglePatents/IFIClaims), ChEMBL Data, PatentsView Data (USPTO), IFI CLAIMS Patent Data Enrichments (with standardized assignee/applicant names), Dimensions.ai (Digital Science), etc.) OntoChem SciWalker Open Data project [BigQuery] Cf. OntoChem SciWalker Open Data - the ontological search framework, https://www.sciwalker.com, see Documentation and  “Expert Search explained” Tab)\
See an example Google BQ report “"FDA Orange Book patents in US Litigation" at p.13 OntoChem  IC-SDV 2019 presentation SciWalker integrating open access & private sources
Wetherbee, I., Arneson, B., Boyer, S., Frommer, J., Mehta, V., 2020. What’s new with scientific content in Google patents. 2020 Revision of presentation at ACS Fall 2019 National Meeting, 33 p. [Introduction of chemical searches in GooglePatents (p.5-17), download chemical data from result, import to BigQuery for analysis in
 Google Data Studio (p.23-24)] [Example: benazepril query gives [4/12/2021] 72,564 results, for which download with Concepts gives a csv file with 2,086,715 lines with entities names, domain, InChI & Smiles and section(s) of a patent document]
Wetherbee, I., Weber, L., Bolton, E., Mehta, V., Boyer, S., Frommer, J., 2020. The Integration of Chemistry with Everything Else. Presented at ACS Fall 2000 National Meeting. 32 p. https://doi.org/10.1021/scimeetings.0c06895 (abstract,  pdf presentation and video, 14:19)
Boyer, S., Wetherbee, I., Böhme, T., Irmer, M., Kruse, K., Püschel, A., Ruttkies, C., Weber, L., 2020 Comprehensive search for compounds and chemical reactions in big query. Presented at ACS Fall 2000 National Meeting, August 2020. 18 p. https://doi.org/10.1021/scimeetings.0c06896 (abstract and pdf presentation)

April 12, 2021
Symposium: Machine Learning and AI for Organic Chemistry: Machine Learning and AI for Organic Chemistry: Session 1
04:00pm - 07:00pm USA / Canada - Eastern - April 12, 2021
Connor Coley, Organizer, Presider (MIT)
Track: [CINF] Division of Chemical Information
Tags: Co-sponsor - Cooperative ORGN: Division of Organic Chemistry


Data-driven design of synthesizable molecules and unprecedented reactions
04:35pm - 05:05pm USA / Canada - Eastern - April 12, 2021
Marwin Segler, Presenter (Westfälische Wilhelms-Universität Münster,Germany; Microsoft Research, Cambridge, UK)
Track: [CINF] Division of Chemical Information
..We will discuss recent developments for enabling the automated design of synthesizable molecules. One possibility is the rapid estimation of synthesizability via distilled synthesizability scores, where the outcome of a synthesis planning algorithm is approximated by machine learning models. Second, we will present a generative model for molecules, which directly outputs synthesis trees.  … application of AI in organic chemistry is the invention and discovery of novel organic methodology, which can enable the synthesis of novel kinds of molecules. Based on graph theoretic consideration, we will present and contextualize a general method for the computer-driven invention of unprecedented reactions.
See also:
Bradshaw, J., Paige, B., Kusner, M.J., Segler, M.H.S., Hernández-Lobato, J.M., 2019. A Model to Search for Synthesizable Molecules. arXiv:1906.05221, 25 p.


Multi-label classification models for the prediction of cross-coupling reaction conditions
06:20pm - 06:50pm USA / Canada - Eastern - April 12, 2021
Michael Maser, Presenter    |     Alexander Cui    |     Serim Ryou    |     Travis DeLano    |     Yisong Yue    |     Sarah Reisman  (California Institute of Technology, Pasadena, CA)
Track: [CINF] Division of Chemical Information
Machine-learned ranking models have been developed for the prediction of substrate-specific cross-coupling reaction conditions. Datasets of published reactions [Reaxys database] were curated for Suzuki [145k reations], Negishi[6.4K] and C–N couplings [36,5k], as well as Pauson–Khand reactions [2.7k]. String, descriptor, and graph encodings were tested as input representations, and models were trained to predict the set of conditions used in a reaction as a binary vector. Unique reagent dictionaries categorized by expert-crafted reaction roles were constructed for each dataset, leading to context-aware predictions. We find that relational graph convolutional networks and gradient-boosting machines are very effective for this learning task, and we disclose a novel reaction-level graph-attention operation in the top-performing model.
See also:
Maser, M.R., Cui, A.Y., Ryou, S., DeLano, T.J., Yue, Y., Reisman, S.E., 2021. Multilabel Classification Models for the Prediction of Cross-Coupling Reaction Conditions. J. Chem. Inf. Model. 61, 156–166. https://doi.org/10.1021/acs.jcim.0c01234  (See Table 1. Statistical Summary of Reaction Data Sets with Reaxys Queries, p.157)

https://acs.digitellinc.com/acs/live/8/page/18/1?eventSearchInput=&eventSearchDate=2021-04-12&eventSearchTrack=83&eventSearchTag=0


April 13, 2021 Tuesday
Symposium
Machine Learning and AI for Organic Chemistry: Machine Learning and AI for Organic Chemistry: Session 2
09:00am - 12:00pm USA / Canada - Pacific - April 13, 2021
Connor Coley, Organizer, Presider (MIT)
Track: [CINF] Division of Chemical Information
Tags: Co-sponsor - Cooperative ORGN: Division of Organic Chemistry


Predicting C-N cross-coupling yields from literature data: Lessons from a machine learning study
12:45pm - 01:05pm USA / Canada - Eastern - April 13, 2021
Martin Fitzner¹, Presenter, Georg Wuitschik¹, Raffael Koller¹, Jean-Michel Adam¹, Torsten Schindler¹, Jean-Louis Reymond²
(1) F Hoffmann-La Roche AG, Basel, Switzerland; (2)  Univ of Bern, Switzerland
Track: [CINF] Division of Chemical Information
…In previous work, we gathered a large amount of Pd-catalyzed cross-coupling reactions from Reaxys, Scifinder and the USPTO, cleaned them and performed statistical analysis. This resulted in cheat sheets, which are a first step to data-driven guides for immediate support of practitioners. In this work, we built on this dataset, which comprises more than 60,000 Buchwald-Hartwig reactions. We applied machine-learning (ML) methods to model the relationship between reactants/reagents and the yield. Three different methods have been investigated: i) multiple fingerprint features; ii) reaction-adapted attention-based graph neural networks; and iii) our custom chemical descriptors with tree-based models. While we find that method iii) consistently outperforms other methods on various data-splits with good metrics, in general the predictive power diminishes drastically when looking at test sets which mimic typical screening experiments. We trace this back to bias and issues with the data diversity, in particular in the space of reagents.
See also
Fitzner, M., Wuitschik, G., J. Koller, R., Adam, J.-M., Schindler, T., Reymond, J.-L., 2020. What can reaction databases teach us about Buchwald–Hartwig cross-couplings? Chemical Science 11, 13085–13093. https://doi.org/10.1039/D0SC04074F (+supporting information)


Learning from literature-extracted synthesis actions for organic synthesis
01:45pm - 02:05pm USA / Canada - Eastern - April 13, 2021
Dr. Alain C. Vaucher, Presenter    |     Philippe Schwaller    |     Joppe Geluykens    |     Federico Zipoli    |     Teodoro Laino (IBM Zurich Research Laboratory, Switzerland)
Track: [CINF] Division of Chemical Information
…Our recent work addresses challenges [of availability of reaction knowledge accumulated in the literature in a machine-friendly format] in the following way. First, we design a transformer-based machine learning model to extract experimental actions from text. This model is pre-trained on a corpus of 2M sentences and associated actions obtained with a rule-based model, and fine-tuned on a set of more than 2000 hand-annotated sentences. Second, the fine-tuned model is applied to experimental procedures from patents to generate a data set of 0.8M chemical equations and associated action sequences, with which, in a third step, we train another machine learning model predicting experimental operations for arbitrary reactions given in SMILES format…
See also:
Vaucher, A.C., Zipoli, F., Geluykens, J., Nair, V.H., Schwaller, P., Laino, T., 2020. Automated extraction of chemical synthesis actions from experimental procedures. Nature Communications 11 (1) , 3601. https://doi.org/10.1038/s41467-020-17266-6
p.3 Synthesis actions. The experimental procedures we consider in this work come from patents and represent single reaction steps. To conduct the full synthesis of a molecule, several such reaction steps are combined.


April 13, 2021
Symposium
Machine Learning in Materials Informatics: Methods and Applications:  
08:00pm - 11:00pm USA / Canada - Eastern - April 13, 2021
Dr Yuling An, Organizer, Presider  (Schrodinger LLC New York)
Track: [CINF] Division of Chemical Information
Tags: Co-sponsor - Nominal POLY: Division of Polymer Chemistry Co-sponsor - Nominal CATL: Division of Catalysis Science and Technology Co-sponsor - Nominal COMP: Division of Computers in Chemistry



Predictive chemical deformulation: A deep generative approach
09:05pm - 09:20pm USA / Canada - Eastern - April 13, 2021
Emre Sevgen, Presenter¹, Edward Kim¹, Brendan Folie¹, Ventura Rivera¹, Jason Koeller¹, Emily Q. Rosenthal², Presenter, Andrea Jacobs², Bryce Meredig¹, Julia Ling¹
(1) Citrine Informatics; (2) CAS
Track: [CINF] Division of Chemical Information
In this work, CAS and Citrine Informatics aim to achieve predictive deformulation using a variational autoencoder method, applied to a subset of CAS’s collection of over 5 million scientist-curated formulated product recipes. With this method, deformulation is achieved with significantly more accuracy than a simpler nearest-neighbor approach.
See also
Citrine Informatics. Better tailor-made formulations, quicker Case Study. February 2020. 3 p.
Citrine Informatics.Solvent Blend Discovery Accelerated (Case Study), April 2020, 4p.
Cf.
Han, R., Yang, Y., Li, X., Ouyang, D., 2018. Predicting oral disintegrating tablet formulations by neural network techniques. Asian Journal of Pharmaceutical Sciences 13, 336–342. https://doi.org/10.1016/j.ajps.2018.01.003

https://acs.digitellinc.com/acs/live/8/page/18/1?eventSearchInput=&eventSearchDate=2021-04-13&eventSearchTrack=83&eventSearchTag=0