《麦肯锡(McKinsey):人工智能在生物制药研究中的应用-集中精力和扩大规模的时机已至(英文版)(11页).pdf》由会员分享,可在线阅读,更多相关《麦肯锡(McKinsey):人工智能在生物制药研究中的应用-集中精力和扩大规模的时机已至(英文版)(11页).pdf(11页珍藏版)》请在三个皮匠报告上搜索。
1、Life Sciences PracticeAI in biopharma research:A time to focus and scaleBy focusing on specific scientific and operational pain points and fully integrating AI into research workflows,biopharma companies can deliver greater patient impact and significant value.October 2022 Janiecbros/Getty ImagesThi
2、s article is a collaborative effort by Alex Devereson,Erwin Idoux,Matej Macak,Navraj Nagra,and Erika Stanzl,representing views from McKinseys Life Sciences Practice.Despite recent advancements,1 biopharma research in drug R&D remains expensive and time-consuming,although there are numerous opportuni
3、ties to build capabilities that enhance productivity and provide probability-of-success gains.In this time of rapid growth of AI in biopharma,attention today is on how to make the most of the opportunity to deliver value at scale by fully integrating AI approaches into scientific process changes.In
4、this article,we outline how biopharma companies can harness AI-driven discovery to deliver patient benefit,and why now is the time for a shift from pursuing select marquee partnerships and self-contained capability builds,to focusing on coordinated investment in research AI with impact to show for i
5、t.The goal of the research phase in drug R&D is to generate as many quality drug candidates as possible,as quickly as possible,with the highest probability of successful transition to clinical development.The discovery process has historically been a convergent,stepped,passfail funnel process with a
6、ttrition at every stepa process that is highly inefficient given the number of compounds initially tested.2 Ideally,this process should only promote compounds for testing that are relevant for targets that would lead to effective drugs for patients.AI can help identify the most promising compounds a
7、nd targets at every step of the value chain so that fewer,more successful experiments are conducted in the lab to achieve the same number of leads.The AI-driven drug discovery industry:Jury still out on impactThe AI-driven drug discovery industry has grown significantly over the past decade,fueled b
8、y new entrants in the market,significant capital investment,and technology maturation.These AI-driven companies fall broadly into two categories:providers of AI enablement for biopharma as a service only,including software as a service(SaaS);and providers of AI enablement that have,in parallel with
9、their services,their own AI-enabled drug development pipeline(see sidebar“Why now is the time for AI-enabled drug discovery”).Our research has identified nearly 270 companies working in the AI-driven drug discovery industry,with more than 50 percent of the companies based in the United States,though
10、 key hubs are emerging in Western Europe and Southeast Asia.3 The number of AI-driven companies with their own pipeline is still relatively small today 1 Fabio Pammolli,et al.,“The endless frontier?The recent increase of R&D productivity in pharmaceuticals,”Journal of Translational Medicine,April 9,
11、2020.2 Less than 0.1 percent of candidate molecules pass from screening to Phase I,with approximately 30 percent of new molecular entity costs being spent in discovery(in the context of$1.1 billion to$2.8 billion spent to bring each drug to market).Oliver J.Wouters et al.,“Estimated research and dev
12、elopment investment needed to bring a new medicine to market,20092018,”JAMA,March 3,2020.Why now is the time for AI-enabled drug discoveryTechnology advances and regulatory openness to innovation have now combined to make AI-enabled drug discovery a practicable proposition.In Europe,regulatory openn
13、ess to in silico and synthetic-derived insights has been facilitated by EU regulation ICH M7 EU,which enables quantitative structure-activity relationship(QSAR)assessment of toxicity instead of traditional assay-based approaches.At the same time,the US Food and Drug Administration(FDA)recognizes sin
14、gle-arm trials in rare diseases with control groups incorporating real-world evidence(RWE).In parallel,technology is advancing on two fronts with(1)standardized approaches to industrialization and scaling of machine learning(ML)for example,MLOps(ML operations)and DataOps(data operations)alongside cu
15、stomized services and platforms;and(2)development of deep-learning approaches for designing new molecules and computer vision,which are increasingly accessible through public code repositories and academic literature.2AI in biopharma research:A time to focus and scale(approximately 15 percent have a
16、n asset in preclinical development).Those with new molecular entities(NMEs)in clinical development(Phase I and II)have predominantly in-licensed assets or have developed assets using traditional techniques.4The growth in the AI-driven drug discovery space has caught the attention of established biop
17、harma companies,and there has been a rapid rise in partnerships between traditional biopharma companies and AI-driven drug companies(Exhibit1).However,there is a significant concentration in partnership activity and funding toward a small number of AI-driven players with high valuations,multiple dea
18、ls,and significant capital raised(62percent CAGR in investment over the past decade).Over half the capital invested in the space is concentrated in only ten companies(all based in the United States or United Kingdom).This is partly because of the difficulty biopharma companies and investors have in
19、evaluating the long tail of AI-driven players.We have seen biopharma companies that are deeply interested in this space struggle to determine what emerging players do,where they operate along the value chain,the distinctiveness of their technology,and which technologies have demonstrable impact.Exhi
20、bit 1Web 2022AiBiopharmaResearchExhibit 1 of 4Investment in emerging AI-driven discovery players increased for a decade before the recent public-market downturn.Companies founded,pharma partnerships by yearIncludes companies founded from previous years.Venture capital.Source:IQVIA Pharma Deals;Pitch
21、book(data has not been reviewed by PitchBook or IQVIA Pharma Deals analysts)002020200Share of funding for top 10 companies with 50%of funding vs others,%Funding by year,$billion51Top 10 companies49OthersFunding and capital investment in AI-driven drug discovery compa
22、niesOtherDebtPre-seedand seedPrivateequityCorporate/M&AIPO/secondaryoferingEarly-stageVCLate-stageVCCompanies foundedPharma partnershipsInvestment in emerging AI-driven discovery players increased for a decade before the recent public-market downturn.3 PitchBook data,2022;IQVIA Pharma Deals,2022.4 P
23、harmaprojects/Informa,2022.3AI in biopharma research:A time to focus and scaleTwo potential obstacles need to be overcome to unlock impact from AI enablement in partnerships among biopharma companies and AI-driven discovery players.First,AI-enabled discovery approaches(including via partnerships)are
24、 often kept at arms length from internal day-to-day R&D;they proceed as an experiment and are not anchored in a biopharma companies scientific and operational processes to achieve impact at scale.Second,investment in digitized drug discovery capabilities and data sets within internal R&D teams is al
25、l too frequently to leverage partner platforms and enrich their IP,rather than building the biopharmas end-to-end tech stack and capabilities.When hurdles are overcome,partnerships can come to fruition,and examples exist across the discovery value chain.AstraZenecas long-standing collaboration with
26、BenevolentAI resulted in the identification of multiple new targets in idiopathic pulmonary fibrosis,with subsequent broadening of the scope to other therapeutic areas(TAs).5 Sumitomo Dainippon Pharma worked with Exscientia to identify DSP-1181 for obsessive compulsive disorder in less than a quarte
27、r of the time typically taken for drug discovery processes(under 12months versus four and a half years)with ambitions to enter the molecule into Phase I trials.6Similarly,building AI-enablement capabilities in-house within biopharma companies is difficult,assembling the cross-functional teams requir
28、ed to drive the transformation is challenging,and it has been observed that AI enablement is often implemented in a relatively isolated way.AI-enabled approaches are often undertaken separately from day-to-day science,with AI-based tools not fully integrated into routine research activities.Biopharm
29、a companies,therefore,need to strike a balance between internal capability building and partnerships with AI-enabled drug discovery companies.Successful biopharma partnerships in the AI space should have some core benefits:biopharma companies gain access to technology(AI platforms,algorithms,and inf
30、rastructure),data(such as curated labeled cell images,screening,ADMET7 data),talent(a ready supply of data scientists and data engineers to build AI pipelines while training biopharma talent),and assurances of data protection in relation to a highly specific strategic intent to maximize patient impa
31、ct(for example,to co-develop a certain molecule class in a specific TA).Substantial impact from building enterprise capabilities in-houseWhen biopharma companies successfully integrate AI processes in day-to-day science and assembles cross-functional teams with the right skill sets(data science,engi
32、neering,software development,epidemiology,discovery sciences,clinical,and design)we have observed significant impact along the value chain(Exhibit 2):Hypothesis generation capabilitiessimplified hypothesis generation tasks in experimental biology fields from several weeks of researcher time to curat
33、ed lists in minutes by combining real-world data(RWD),genomics data,and scientific literature through a knowledge graph for target identification Large-molecule-structure inference 100times acceleration in time to generation of protein structures(for example,for peptide or mRNA-vaccine-antigen gener
34、ation)for target identification Computer vision technologyup to ten times acceleration achieved for screening-plate-image analysis,with higher accuracy than classical approaches,harnessing deep-leaning approaches(for instance,convolutional neural networks)for target validation and hit identification
35、5“BenevolentAI announces 3-year collaboration expansion with AstraZeneca focused on systemic lupus erythematosus and heart failure,”BenevolentAI press release,January 13,2022.6“Sumitomo Dainippon Pharma and Exscientia joint development new drug candidate created using artificial intelligence(AI)begi
36、ns clinical trial,Exscientia press release,”January 30,2020.7 ADMET refers to chemical absorption,distribution,metabolism,excretion,and toxicity.4AI in biopharma research:A time to focus and scale In silico medicinal chemistry30 to 50 percent acceleration in small molecule,high-throughput screening,
37、using approaches such as molecular property prediction in an iterative screening loop(versus the existing approach of randomized selection of compounds)for hit identification In silico chemi-informaticsmore than two times improvement over baseline on the key metric of“efficacy observed,”over 100 tim
38、es the number of in silico experiments possible compared with previous screening,and faster time for design of compounds for optimization of drug delivery efficacy for lead optimizationExhibit 2Web 2022AiBiopharmaResearchExhibit 2 of 4AI has already delivered value across the research value chain.Ex
39、amples of AI-driven innovation in biotech and biopharma and observed impact across the research value chainTarget identifcationTarget validationHit identifcationLead generation/optimizationPreclinicalInsights from data sources(internal and from vendors)to generate novel target hypothesisGene network
40、,biochemical pathway,and cellular-response data integration in target identifcationIn silico,phenotypic,cellular models validate targets/identify biomarkersDisease causality determined within patient groups with signifcant unmet medical needAutomated image analysis for cellular assays through comput
41、er vision technologyMolecular property prediction(virtual screening)Molecular structure and property prediction(eg,protein binding,logP,toxicity)for novel target proteinsRapid design iteration,across small and large molecules,using eg,Generative Adversarial NetworksSafety issue and DMPK prediction u
42、sing internal and public dataHypothesis-driven dosages for adaptive trials and targeted populationsBiopharma unlocked all-inclusive view of complex indication by attributing disease causality through linkages between genomic data and patient electronic medical records(EMRs)Biopharma internalized Alp
43、haFold2 and ColabFold togenerate 3-D models of almost any known,synthesized protein andproteinprotein interactions,reducing access to 3-D structures from 6 months to a few hoursBiotech saw signifcant acceleration of high-throughput screening(HTS)phase(time to 75%hits detectedreduced by 50%)with plat
44、form-based“compound prioritization”algorithmBiopharma leveraged generative machine learning model toexpand library/optimize promising compounds and predict compound efcacy,signifcantly increasing efciency of library expansion,with 60,000 new compounds generatedBiopharma utilized predictive algorithm
45、s to maximize probability of successful PK predictions with 83%of drug development projects progressing to clinic with no PK issuesExamples of AI-driven accelerationExamples from industry and observed impactDrug metabolism and pharmacokinetics.Pharmacokinetic.AI has already delivered value across th
46、e research value chain.5AI in biopharma research:A time to focus and scale Knowledge-graph-based hypothesis generation and drug repurposingrapid identification of novel indications for existing investigational new drugs(INDs)or marketed drugs via genomic information and pathways associated with spec
47、ific disease phenotypes,accelerating time to new treatments for patients,as part of the preclinical phase of R&D Indication finding leveraging genomicsprioritizing indications to pursue for novel mechanisms of action(MoAs),finding new greenfield indications for life cycle management,prioritizing or
48、deprioritizing ongoing programs within clinical plan by stopping low probability of success programs early and reducing patient burden in clinical trials;informing diligence of molecules for licensing with an independent view of biological potential,as part of the preclinical phase of R&D Biopharma
49、companies that maximize the impact of AI enablement can move beyond minimum viable product(MVP)individual use cases and build research systems(Exhibit 3).Exhibit 3Web 2022AiBiopharmaResearchExhibit 3 of 4Research systems harness synergies when AI technologies are embedded into wet lab scientifc proc
50、esses.Parts of a high-throughput screening(HTS)process embedded with AI technology1.High-throughput screen commenced with diverse compound setsScientist selects diverse compound sets(a set of chemical compounds with a wide range of chemical structures)as frst high-throughput screen3.Computer-vision-
51、based hit selectionCell response to each compound is measured using microscope analysis(eg,through computer vision techniques);promising compounds are labeled“hits”2.Automated compound selection and transferUsing HTS machinery,individual compounds are transferred to individual wells of cells under e
52、xperimental conditions4.Automated machine learning(ML)model training from screen outcomesInformation from HTS for frst few plates is automatically transferred into an ML pipeline,which“learns”how cells respond to each kind of chemical structure6.Automated compound selection based on ML recommendatio
53、nsML recommendations are automatically queued and used in the next round of HTS.The cycle continues,with the algorithm continuously learningfrom“real world”outputs.Recommendations triggerscientists to explore new chemical space and begin downstream screening processes more quickly.These recommendati
54、ons feed into the selection of chemical compounds in step 15.Compound library inferencing and prioritizationML algorithm then scans the remainder of the library compounds and predicts which plates should be prioritized to identify the highest number of hits in the next screen123456In silico/on-the-c
55、hip simulationsIn-vitro/wet lab experimentsResearch systems harness synergies when AI technologies are embedded into wet lab scientific processes.6AI in biopharma research:A time to focus and scaleResearch systems harness synergies created by putting AI at the center of the research engine to enhanc
56、e the outcome of experimentsinstead of simply being a preparatory step for real-world experiments in isolation.They act as feedback loops to refine the predictive capability and stability of AI algorithms and inform experimental design(for more key definitions,see sidebar“Glossary of key pharma AI R
57、&D terms”).An example is“iterative screening”:results of an initial round of high-throughput screening8 are used to train a machine learning(ML)algorithm.The ML algorithm can learn which underlying compound structures are most effective against a target and suggest other molecules in the library to
58、prioritize for testing.As the ML algorithm gathers more data,its predictions rapidly become more accurate,and a disproportionately large number of“hits”are identified for the relative amount of the library screened.These research systems reduce overall costs,have higher probability of success,accele
59、rate R&D processes(and therefore time to patient impact),and are fully integrated for specific use cases.What does it take to successfully implement AI in biopharma research?By implementing digital and data science tools and concepts,biopharma can capture the full value of current portfolios and dev
60、elop core technologies,competences,and IP to drive future research(such as AI-enabled large-molecule and antibody design).Current AI-driven drug discovery companies are already developing their own,significantly Glossary of key pharma AI R&D terms Artificial intelligence/machine learningOften used i
61、nterchangeably,AI and ML have subtly different meanings.A subset of AI,machine learning refers to algorithms that improve by design through the addition of new data.AI describes,more broadly,a computer system that can solve problems based on data available in its environment,in the context of achiev
62、ing specific goals.Deep learningA subset of machine learning,deep learning utilizes an artificial neural-network structure,typically involving a very large number of parameters to be calibrated,to identify patterns in large data sets and subsequently to make predictions based on this calibrated neur
63、al network.Molecular property prediction/iterative screeningAn approach at the intersection of machine learning and drug discovery that utilizes machine learning approaches to determine patterns in candidate molecule structures that predict their likelihood of being a successful drug candidate.Typic
64、ally,in an“iterative screening”setup,an ML algorithm will be trained based on historic screening experiments and will predict which of the remaining compounds in the screening library are likely to be hits;as further batches of experiments are conducted,the algorithm increases its precision and pred
65、ictive power.The algorithms learn from both successful and unsuccessful predictions.Knowledge graphsA type of database that integrates data from multiple sources and creates links across them,typically in the same domain(for example,drug discovery).Data is organized in accordance with an ontologya w
66、eb of concepts,including the codified relationships between conceptsand its controlled vocabulary.Knowledge graphs can be used to infer connections across multiple data sources,including link prediction and insight generation.DataOps/MLOpsDataOps(data operations)is an automated,process-oriented meth
67、odology used by analytics and data teams to improve quality and reduce the cycle time of advanced analytics.MLOps(ML operations)refers to software engineering practices applied to IT operations(for example,packaging and deploying production software)in the context of machine learning and artificial
68、intelligence.8 High-throughput screening involves thousands to millions of candidate molecules in a compound library being tested against a specific assay and determined to be effective against a target,or a“hit.”7AI in biopharma research:A time to focus and scalemore cost-efficient drug discovery p
69、ipelines,so it would be beneficial for established players to identify how they,too,can fully integrate novel technologies into standard research processes.While partnering is one optionwhere it provides access to data,technology,and talent,and the risk of partners exploiting a companys IP to become
70、 a future competitor in the medium to long term is lowmarquee partnerships cannot be the only way to develop in-house drug discovery capabilities.As such,it is critical for biopharma companies to work out how to shift from investing in nonintegrated,lighthouse use cases or partnerships to making AI
71、an integral part of everyday research.With this in mind,here are four areas to consider:1.Strategy and design-backed road-mapping.Biopharma companies can develop a top-down,C-level strategy,setting out the ways in which AI-enabled discovery will be a critical enabler of future performance.A signific
72、ant aspect is to understand where the current organizational pain points lie,what the potential gains could be,and where the organization wants to lead the industry(versus only being competitive)in the context of how the space/competitors are expected to move in the future.This strategy needs to be
73、specific,time-bound,linked to value at stake,and have strong alignment among(and sponsorship from)senior leadersincluding the heads of R&D,research,and data science.Underpinning this strategy is the need for sufficient resources(balanced across talent,data,and infrastructure investment)to support th
74、e capability building and talent acquisition required to make it a reality,or recognition of the trade-offs on IP and capability building if only pursuing external partnerships.Alignment between R&D and digital functions is paramount to ensure balanced co-investment(financial and management time)and
75、 for the impact generated from initiatives to be shared appropriately.In addition,it is important to carefully consider which elements of the AI-enabled drug discovery approach will be supported by partnerships versus built in-house.We recommend a design thinking approach to determine which parts of
76、 discovery research to tackle,and in which order.This involves studying,end-to-end,common research processes,where there may be two to three steps that are bottlenecks for researchers,and which could be significantly unlocked via AIfor example,automated image analysis for critical cell assays or lea
77、d optimization.Design thinking could help companies determine which areas could benefit most from AI,the implementation road map,and the success indicators to track progress and impact(for example,time from target identification to candidate selection,costs associated with target identification).For
78、 R&D and data science leaders,the focus should not be solely on advanced-analytics use cases:there is significant value in cracking established problems,with applications such as basic Biopharma companies can develop a top-down,C-level strategy,setting out the ways in which AI-enabled discovery will
79、 be a critical enabler of future performance.8AI in biopharma research:A time to focus and scaleautomation using data transformation pipelines(such as dose response curve fitting),digital operational dashboards,or building data platforms and infrastructure(such as knowledge graphs).For example,build
80、ing a single data platform for all preclinical data generated can prevent experimental duplication and enhance data sharing across the organizationour experience shows this can reduce months of hypothesis generation time to a few days.The impact includes dramatically increased speed,freeing up peopl
81、e for more productive tasks,and increasing quality of analyses.2.Relentless value delivery focused on quarterly value releases(QVRs).It is critical that R&D,data science,and data engineering collaborate closely and iterate on delivery of use cases in an agile way.The research process frequently incl
82、udes specific constraints and ways of working(such as steps and hand offs in the experimental methodology)that need to be accounted for to ensure uptake of the tools and systems that are built(in addition to updating scientific processes and standard operating procedures and introducing financial an
83、d performance-based incentives).To consider AI-enablement delivery holistically,leaders can line up key building blocks,as in this specific example focused on“high-throughput screening”:Blueprinting.Develop a list of use cases across the value chain,prioritizing according to impact,complexity,and bu
84、siness value;then select the highest-need use cases.Digital and analytics solutions.Build and automate screening algorithms that link molecular descriptors(for example,molecule structure in the form of a SMILES9 string)with desired output,or a hit.Data continuum.Collect experimental data in a reusab
85、le way(for instance,with FAIR-data principles10);build master tables from equipment and existing libraries.Tech capabilities.Design and build technical infrastructure and data architecture for data extraction and automated gathering.Talent and agile operating model.Coach data science,data engineerin
86、g,and translator/product owners on tools and delivery methodologies,iteratively testing and learning to deliver products via a collaborative environment.Adoption and scaling(including change management).Design new screening protocols and experimental strategy,incorporating ML-based algorithms.Ensure
87、 the whole research organization(from leaders to lab technicians)understands what the company is trying to achieve and how daily activities need to change.Once key AI-enabled use cases are aligned,delivery must be highly organized so as to demonstrate ongoing impact;core requirements and potential s
88、ynergies must be identified and gaps in ongoing cross-cutting road maps identified.This means departing from long-term road maps delivering impact in multiyear cycles to focus on QVRs(which produce measurable value after each quarterly sprint,such as AI-enablement of a scientific process)while conti
89、nuously reprioritizing based on organizational needs.This approach enables AI use-case development to be built more efficientlyby dynamically front-loading priority data ingestion and team capacitywith mission-critical assets deployed as required(Exhibit 4).All core digital processes in research can
90、 be delivered with incremental quarterly delivery;however,the nature of“value”delivery may vary.Moonshot programs(in tech,this could be the advent of AlphaFold11)require long-term road maps and typically a dedicated ML research group to deliver potentially groundbreaking discoveries with impact in b
91、iopharma.Such programs may not deliver an AI product every quarter such as other digital 9 SMILES refers to simplified molecular-input line-entry system.10 FAIR refers to findable,accessible,interoperable,and reusable.11 AlphaFold is an AI system developed by DeepMind that predicts a proteins 3-D st
92、ructure from its amino-acid sequence.9AI in biopharma research:A time to focus and scaleinitiatives,but an insight,report,or decision should still be delivered on a regular basis.3.IP,capability building,and developing translation expertise through partnerships.While there is certainly evidence for
93、the benefits of partnership in specific areas,including to access unique technologies,data,or solution types,managing these partnerships exclusively at arms length and keeping novel methods or solutions separate from day-to-day research mean that necessary future capabilities for a transformation in
94、 drug discovery may not be built.Biopharma companies should be selective and specific about the capabilities to be delivered by partnerships versus those built in-house.Similarly,a balanced approach to in-house and external talent(notably,the data scientists and data engineers needed to work with re
95、searchers in developing the algorithms and technology backbones to support prioritized areas)is vital.Often overlooked but mission critical for AI enablement,are“translators”or“product owners”with deep business,clinical,scientific,and AI/ML and systems architecture understanding.These profiles have
96、a product ownership mindset and understand and dynamically evaluate all elements of the analytics team to maintain focus on value and impact delivery,thereby assuring successful project delivery.4.Industrialization of AI with MLOps and reusable analytical assets.For the capabilities a biopharma comp
97、any builds in-house,it is essential to have the right enablers in place to support scaling across research activities:the right technology Exhibit 4Web 2022AiBiopharmaResearchExhibit 4 of 4Moving from parallel processes to quarterly value releases touching all six building blocks of an analytics bui
98、ld with links to impact.Parallel programs of foundational eforts,with minimal linkages between a subset of 6“building blocks”of analytics builds,where not all blocks may be considered;no explicit link to business/scientifc fows and impactProgram links all horizontals via releases,touching all 6“buil
99、ding blocks”of an analytics build;links from QVR to impact through scientifc/business fowsDigital/analyticsDataTechnical capabilityCapability building/agileAdoption and scalingValue release happens quarterlyIllustration of holistic AI-enablement value delivery focused on quarterly value releases(QVR
100、s)Nonoptimized AI-enablement deliveryOptimized digital deliveryBlueprintBlueprintDigital/analyticsDataTechnical capabilityValue release mostly at the end of each horizontalQVR1QVR2QVR3Digital deliveryCapability building/agileAdoption and scalingMoving from parallel processes to quarterly value relea
101、ses touching all six building blocks of an analytics build with links to impact.10AI in biopharma research:A time to focus and scaleDesigned by McKinsey Global PublishingCopyright 2022 McKinsey&Company.All rights reserved.Alex Devereson is a partner in McKinseys London office,where Matej Macak is an
102、 associate partner and Navraj Nagra is a consultant;Erwin Idoux is a consultant in the Paris office;and Erika Stanzl is an associate partner in the Zurich office.The authors wish to thank Jeffrey Algazy,Christoforos Anagnostopoulos,Joachim Bleys,David Champagne,Thomas Devenyns,Bryan Lee,Rachel Moss,
103、Michael Steinmann,and Lieven Van der Veken for their contributions to this article.Scan Download PersonalizeFind more content like this on the McKinsey Insights Appinfrastructure and methodologies,especially DataOps and MLOps and an appropriate data architecture(for example,graph databases or Data V
104、ault 2.0 technology).DataOps(data operations)enables companies to gain more value from their data by accelerating the process of building models.MLOps involves ensuring the right platforms,tools,services,and roles with the right team operating model and standards for delivering AI reliably and at sc
105、ale.Technical-architecture enablers to support processing compute-intensive workflows such as AlphaFold,molecular-dynamics simulations,optimization models,and image-recognition workflows are a core requirement.Furthermore,enabling concepts such as Data Vault 2.0 techniques and graph databases are ta
106、ble stakes as AI capabilities scale.To successfully deploy research systems,development teams must build multiple interrelated components(data connectors and pipelines,models,APIs,and visual interfaces)that work seamlessly to drive adoption among end users.Fragmentation of code bases and components,
107、and reduced productivity due to integration challenges,are natural risks that arise when multiple tools are deployed across different domains and teams.Ensuring coding standards in development and harmonization of coding approaches across teams increases long-term productivity and solution robustnes
108、s.Additionally,harmonization enables sharing of reusable components(data connectors,feature libraries,model-based embeddings)across projects:for example,using graph neural-network molecular embeddings for hit prediction and lead optimization for toxicity reduction.As the emerging research platform g
109、rows in complexity,“assetization”of reusable components becomes an increasingly important source of development productivity(with twice the productivity for teams that embrace it)and an important in-house capability that requires a dedicated team with a product-centered mindset.12The question today
110、is whether biopharma companies will move analytics investments beyond a focus on individual projects and marquee partnerships to transforming research at scale.A shift to focusing on specific scientific and operational pain points and building AI into fully integrated research systemswith a road map
111、 to scalewill enable biopharma companies to capture real business and patient impact from using AI in research.12 Industrialization of AI with MLOps is one of 14 foremost trends identified by the McKinsey Technology Council.For more information,see Michael Chui,Roger Roberts,and Lareina Yee,“McKinsey Technology Trends Outlook 2022,”August 24,2022.11AI in biopharma research:A time to focus and scale