Skip to content
Nobel Jews
Portrait of Michael Levitt
Photo: Bengt Nyman from Vaxholm, Sweden, Flickr: IMG_7539 · CC BY 2.0 via Wikimedia Commons

Nobel Prize in Chemistry · 2013

Michael Levitt

South African-born biophysicist who helped turn the computer into a chemistry lab, simulating how proteins fold, move and work.

The Nobel citation: “for the development of multiscale models for complex chemical systems”
Born
May 9, 1947, Pretoria, South Africa
Affiliation at the time
Stanford University School of Medicine, USA

Chemistry prize

2013

Shared with 2 other laureates.

Age that year

66years

Born in 1947.

Headline credited impact

45,000–180,000people benefited

People treated with monoclonal antibody medicines worldwide (share via antibody humanization modeling). How it was built

Sources cited

29

Fact-checked September 24, 2026.

  • He finished his last two years of high school in a single summer and started at the University of Pretoria a few months before turning 16.
  • To win a PhD place at Cambridge, the shy student stopped Nobel laureate Max Perutz in a corridor, wearing the suit from his bar mitzvah.
  • His first protein calculations ran on Golem A, a home-built Weizmann Institute computer with about 300,000 bytes of memory, fed with punched cards.
  • For two years in Cambridge he worked at the lab only in the mornings and cared for his baby son at home, telecommuting before the word existed.
  • He holds British and Israeli citizenship, and all three of his sons did their army service in Israel.

The breakthrough

Computer models of proteins, from single atoms to whole molecular machines

Proteins are chains of amino acids that fold into precise 3D shapes. Their atoms push and pull on one another, and those forces decide both a protein's shape and how it moves. In 1967, aged 20, Levitt went to the Weizmann Institute in Israel and, with Arieh Warshel, wrote the computer program that could calculate those forces for any molecule. He then tried something new: running the calculation on whole proteins, lysozyme and myoglobin, and letting the computer ease every atom toward a more comfortable position.

Two big ideas followed, both with Warshel. The first was simplification. In 1975 they treated each amino acid as one or two beads instead of about ten atoms, which made it possible to simulate a small protein folding up. The second was mixing two kinds of physics. In 1976 they modeled the enzyme lysozyme using quantum physics only for the few atoms where a chemical bond breaks, and simpler classical physics for the rest. Think of a video game that draws the scene near the player in full detail but shows distant hills as rough shapes.

Later, Levitt showed that surrounding a simulated protein with thousands of water molecules makes it behave far more realistically. Together these ideas, called multiscale modeling, earned the 2013 Nobel Prize in Chemistry, shared with Warshel and Martin Karplus.[2],[3],[4],[5],[6],[17],[18],[20]

“Moving experimental chemistry into cyberspace should be of clear importance to biomedical science, as it allows one to accelerate the testing of hypotheses.”
Michael Levitt, From his written Nobel lecture (2013), explaining why computer simulation matters for medicine.[6]

What it meant for humanity

Almost everything a living cell does is carried out by proteins, molecules of thousands of atoms that fold into exact shapes and move as they work. Experiments can capture those shapes but struggle to show the motion. Levitt spent five decades writing programs that fill the gap.

His early papers set patterns still in use. In 1969 he and Shneior Lifson made the first energy calculation that refined a whole protein structure. In 1975, with Arieh Warshel, he folded a simplified protein on a computer, and in 1976 the pair published the first general model mixing quantum and classical physics to follow an enzyme reaction. A 2009 review called such hybrid methods the method of choice for modeling reactions in biological molecules. In 1988 he and Ruth Sharon showed that simulating a protein in a box of water gives far more realistic results, and he later wrote that almost all simulations now do this.

One application reached patients. In 1987 the start-up Protein Design Labs asked him to build 3D models of antibodies. Mouse antibodies can make good drugs, but the human immune system attacks them. Cary Queen used Levitt's models to decide which mouse building blocks to keep when rebuilding an antibody on a human frame. Their 1989 antibody became daclizumab, in 1997 the first humanized antibody approved by the US FDA. The company's humanization patents were later licensed for Herceptin, Avastin, Xolair, Lucentis, Tysabri and Actemra, medicines used for cancer, asthma, multiple sclerosis and arthritis, among other diseases.

The Nobel committee noted that the same modeling helps chemists improve catalysts, drugs and solar cells.

  • With Shneior Lifson he carried out the first energy minimization of an entire protein structure, published in 1969.[5],[19]
  • The 1976 Warshel-Levitt lysozyme paper, the first general hybrid quantum-classical model of an enzyme reaction, has more than 4,100 citations in Crossref's records.[5],[21],[22]
  • His 1988 simulation with Ruth Sharon showed a protein in water stays much closer to its real structure than one in a vacuum; he wrote that almost all simulations now include thousands of water molecules.[6],[20]
  • His antibody models helped Protein Design Labs build the humanized antibody that became daclizumab, the first humanized antibody drug approved by the US FDA, in 1997.[6],[12],[14],[26]
  • Protein Design Labs' humanization patents were licensed for Herceptin, Avastin, Xolair, Lucentis, Tysabri and Actemra. Actemra alone had treated more than one million people by 2019.[13],[14],[16]
  • Between 1986 and 2013 he mentored 14 PhD students and 29 postdoctoral fellows. In 2017 he co-wrote a PNAS study showing US research funding shifting away from scientists younger than 46.[6],[27]

Impact in numbers

Most of Levitt's influence cannot honestly be counted. Simplified bead models, hybrid quantum-classical models and simulations of proteins in water are now everyday tools in chemistry, structural biology and drug research. But their value is mixed with that of experiments, faster computers and thousands of other scientists, including his co-laureates, so, as in Arieh Warshel's profile, no number is claimed for them. One pathway is concrete enough to estimate: his antibody models fed into Protein Design Labs' humanization method, whose patents were licensed for several widely used antibody medicines. That single low-confidence claim gives him a very small share (0.3 percent) of the people treated with antibody medicines. Beyond medicine, he helped build computational biology into a field, trained dozens of scientists and has campaigned, with data, for young scientists to get the chances his generation had.

Fundamental scienceHealthTechnology

Each number is the laureate’s credited share of a real-world outcome, cumulative to 2025. The whole outcome, the share of credit, and the reasoning are shown so you can check the arithmetic. Outcomes shared with other laureates are counted once on the impact page.

  • Low confidenceRippleModeledHealth

    People treated with monoclonal antibody medicines worldwide (share via antibody humanization modeling)

    45,000–180,000

    people benefited, credited share

    That is 0.3% of 15–60 million people benefited since 1986.

    How this number was built

    Whole outcome reused from the Milstein profile so totals line up: 15M-60M people treated with antibody medicines. Cross-check: Roche reports 2.7M rituximab patients by 2017; rituximab was $6.8B of a $115.2B market in 2018 (5.9%), implying about 46M, inside the range. Share 0.003 = 0.2 x 0.15 x 0.1. Drugs licensed under PDL's humanization patents were about 15% of 2018 sales (Herceptin, Avastin, Xolair: $16.8B of $115.2B = 14.6%), rounded up to 0.2 for Lucentis, Actemra, Tysabri, Perjeta, Kadcyla and others. Humanization gets about 0.15 of the credit for those drugs; hybridomas, target biology and development get the rest. Levitt's structural models get about 0.1 of PDL's humanization work, led by Cary Queen and built on Greg Winter's CDR grafting. Patients are assumed proportional to sales, a crude step. Result: about 45,000-180,000 people treated, not cured.[6],[12],[13],[14],[15],[16]

    Sources: NobelPrize.org (Nobel Foundation); PubMed (National Library of Medicine); US Securities and Exchange Commission (EDGAR); Journal of Biomedical Science (via PubMed Central); Roche; Roche

The double edge

No harm has been traced to Levitt's modeling methods, which are research tools. The main controversy is his public role in the COVID-19 pandemic. Speaking as a Nobel laureate but outside his field, he made confident forecasts that proved badly wrong, argued against lockdowns and signed the Great Barrington Declaration, which proposed letting low-risk people build immunity through infection. Epidemiologists criticized his forecasts, and a fellow Nobel laureate said his prize gave his policy views undue influence. He later admitted the forecasts were wrong and said he should have given a range instead of a single number. A smaller issue touches his antibody work: a later version of daclizumab, the drug that grew out of the 1989 humanized antibody he co-authored, was withdrawn in 2018 after regulators found it could cause serious and potentially fatal immune reactions.

  • Moderate

    Wrong COVID-19 forecasts from a Nobel platform

    By his own account, in March 2020 he told Israel's prime minister he would be surprised if the country saw more than 10 COVID-19 deaths; the count passed 6,400 by May 2021. On 25 July 2020 he tweeted that the US outbreak would be over within four weeks with fewer than 170,000 deaths. US death certificates list about 190,000 COVID-19 deaths by the end of August 2020 and about 386,000 for the year. He later said he felt really bad about declaring it in such a combative way.[8],[9],[10]

  • Moderate

    Campaigning against lockdowns

    He signed the Great Barrington Declaration of October 2020, which proposed letting people at low risk of death build immunity through natural infection, and told a roundtable hosted by Florida's governor that exposing young people made sense. The WHO's director-general called such herd-immunity strategies unethical. Critics, including Nobel laureate Randy Schekman, said Levitt's prize gave his policy views undue influence.[8],[11]

  • Minor

    A daclizumab medicine withdrawn for safety

    Daclizumab grew out of the 1989 humanized antibody paper on which Levitt was a co-author. A later version, daclizumab beta (Zinbryta), authorised in the EU in 2016 for multiple sclerosis, was withdrawn in 2018, and regulators confirmed it could cause serious and potentially fatal immune reactions in the brain, liver and other organs. Levitt's role in the original design was small and indirect.[12],[14],[28]

Against the odds

Levitt was born in Pretoria in 1947 into a Jewish family with roots in Lithuania, the homeland of most South African Jews. The generation before him had met real hostility. As antisemitism grew, a 1930 Quota Act, aimed in practice at Jews, cut immigration from Lithuania, Poland and other Eastern European countries. A 1937 Aliens Act shut out most Jews fleeing Nazi Germany. In 1948 the National Party, whose Afrikaner nationalist base had used antisemitism to build support, took power. In his father's birthplace, Plungė, about 1,800 Jews were shot in 1941 after the German invasion, with local Lithuanian collaborators taking part. The sources do not say whether any of his relatives were among them.

Levitt himself grew up comfortable and, by his own account, happy. Under apartheid, antisemitism was no longer a major public issue, and he belonged to the privileged white minority. His obstacles were of other kinds. Cambridge at first turned him down, a reviewer mocked his thesis methods as worth an F, and computer biology was met with deep skepticism. After he moved to Israel, war followed: he arrived two months before the 1973 Yom Kippur War, and in 1991 he had to leave his family in Israel while missiles were falling. In 1987 some Weizmann Institute colleagues pushed him to leave, and he built the rest of his career at Stanford.

  • 1941

    Other

    Before he was born, the Jewish community of his father's hometown, Plungė in Lithuania, was destroyed: about 1,800 Jews were killed in 1941, according to a survivor's memoir. No source links his relatives to the massacre.[25]

  • 1967

    Other

    John Kendrew first told him the Cambridge laboratory had no space and could not even consider him. He won a place only after asking again and confronting Max Perutz in person.[2]

  • 1972

    Other

    Computational structural biology was new and met great skepticism. A reviewer ridiculed his thesis methods for computing forces on molecules, saying they would have earned an F as an undergraduate term paper; most thesis chapters were never published.[2]

  • 1973

    War

    Arrived at the Weizmann Institute two months before the October 1973 war. His Israeli colleagues, including Warshel, were called up, and he became deeply depressed; talking about work with Warshel afterward helped them both.[2]

  • 1987

    Other

    Some members of his Weizmann Institute department acted to encourage him to leave. He took leave for Stanford, and when he later sought a joint appointment, the same people objected. The Times of Israel wrote that he left because he felt professionally stymied.[2],[7]

  • 1991

    War

    At the start of the 1991 Gulf War he had to return to Stanford to teach while missiles were falling on Israel and his family there feared a nerve gas attack.[2]

Jewish background

Both parents JewishIdentified as Jewish, secular

Levitt was born in Pretoria into a Jewish family with roots in Lithuania, the homeland of most South African Jews. In his Nobel autobiography he writes that his father was born in Plungė, Lithuania, and came to South Africa at 10, and that his mother was born in Johannesburg to parents from Czechoslovakia; he mentions his bar mitzvah. Reference works describe the family as Jewish but say little about his mother's own ancestry. He married Rina in Israel in 1968, moved to Israel in 1979 and taught at the Weizmann Institute until 1987. He holds British and Israeli citizenship, and his three sons served in the Israeli army.[2],[7],[8],[23],[24],[29]

Key dates

  1. May 9, 1947

    Born in Pretoria, South Africa, to a Jewish family with roots in Lithuania.[1],[2],[29]

  2. 1963

    Passes his school-leaving exams early and enters the University of Pretoria to study applied mathematics, a few months before turning 16.[2]

  3. November 18, 1963

    Leaves for London to visit his scientist uncle and aunt and decides to stay; his mother, sister and brother follow.[2]

  4. October 13, 1967

    After a physics degree at King's College London, arrives in Israel to work with Shneior Lifson and Arieh Warshel at the Weizmann Institute.[2],[6],[7]

  5. August 1968

    Marries Rina, whom he met in Rehovot, in Israel, then starts his PhD at the MRC Laboratory of Molecular Biology in Cambridge.[2]

  6. 1969

    With Lifson, publishes the first energy minimization of an entire protein structure.[5],[19]

  7. 1971

    Completes his PhD at the University of Cambridge.[2],[27]

  8. February 1975

    With Warshel, publishes an early landmark computer simulation of protein folding, using a simplified bead model.[5],[17]

  9. 1976

    With Warshel, publishes the general quantum-classical (QM/MM) model of the enzyme lysozyme's reaction.[5],[18]

  10. September 1979

    Moves to Israel for a permanent post at the Weizmann Institute; in 1980 he becomes chair of its Department of Chemical Physics.[2],[7]

  11. July 1987

    Joins Stanford University and begins consulting for Protein Design Labs on antibody models.[2],[6]

  12. December 1989

    Co-authors the Protein Design Labs paper on a humanized antibody against the interleukin-2 receptor, designed with help from a computer model.[12]

  13. October 9, 2013

    Awarded the Nobel Prize in Chemistry with Martin Karplus and Arieh Warshel for multiscale models of complex chemical systems.[3]

  14. 2020

    Turns to COVID-19 data; his public forecasts of a quick end to the pandemic prove wrong and draw wide criticism.[8],[9]

Sources

  1. 1.Michael Levitt - Facts · NobelPrize.org (Nobel Prize Outreach), 2013
  2. 2.Michael Levitt - Biographical · NobelPrize.org (Nobel Foundation), 2013
  3. 3.The Nobel Prize in Chemistry 2013 - Press release · Royal Swedish Academy of Sciences via NobelPrize.org, 2013
  4. 4.Taking the experiment to cyberspace (Popular information, Nobel Prize in Chemistry 2013) · Royal Swedish Academy of Sciences via NobelPrize.org, 2013
  5. 5.Scientific Background: Development of multiscale models for complex chemical systems · Royal Swedish Academy of Sciences via NobelPrize.org, 2013
  6. 6.Birth and Future of Multiscale Modeling for Macromolecular Systems (Nobel Lecture, 8 December 2013) · NobelPrize.org (Nobel Foundation), 2013
  7. 7.3 Jewish professors -- two of them Israeli -- share 2013 Nobel Prize in chemistry · The Times of Israel, 2013
  8. 8.He's a Stanford professor and a Nobel laureate. Critics say he was dangerously misleading on Covid · STAT News, 2021
  9. 9.Prof Michael Levitt: here's what I got wrong · UnHerd, 2020
  10. 10.Provisional COVID-19 Death Counts by Week Ending Date and State (NCHS) · US Centers for Disease Control and Prevention, National Center for Health Statistics, 2026
  11. 11.WHO chief says herd immunity approach to pandemic 'unethical' · The Guardian, 2020
  12. 12.A humanized antibody that binds to the interleukin 2 receptor (Queen C, ... Levitt M, ... Waldmann TA), PNAS 86:10029-10033 · PubMed (National Library of Medicine), 1989
  13. 13.PDL BioPharma, Inc. Annual Report (Form 10-K) for the year ended December 31, 2013 · US Securities and Exchange Commission (EDGAR), 2014
  14. 14.Development of therapeutic antibodies for the treatment of diseases (Lu RM et al.), Journal of Biomedical Science 27:1 · Journal of Biomedical Science (via PubMed Central), 2020
  15. 15.FDA approves Rituxan Hycela (rituximab and hyaluronidase human) for subcutaneous injection in certain blood cancers · Roche, 2017
  16. 16.FDA approves Roche's Rituxan (rituximab) in children with two rare blood vessel disorders · Roche, 2019
  17. 17.Computer simulation of protein folding (Levitt M, Warshel A), Nature 253:694-698 · PubMed (National Library of Medicine), 1975
  18. 18.Theoretical studies of enzymic reactions: dielectric, electrostatic and steric stabilization of the carbonium ion in the reaction of lysozyme (Warshel A, Levitt M), J Mol Biol 103:227-249 · PubMed (National Library of Medicine), 1976
  19. 19.Refinement of protein conformations using a macromolecular energy minimization procedure (Levitt M, Lifson S), J Mol Biol 46:269-279 · PubMed (National Library of Medicine), 1969
  20. 20.Accurate simulation of protein dynamics in solution (Levitt M, Sharon R), PNAS 85:7557-7561 · PubMed (National Library of Medicine), 1988
  21. 21.Crossref metadata record and citation count for Warshel and Levitt (1976), J Mol Biol 103:227-249 · Crossref, 2026
  22. 22.QM/MM methods for biomolecular systems (Senn HM, Thiel W), Angew Chem Int Ed 48:1198-1229 · PubMed (National Library of Medicine), 2009
  23. 23.South Africa Virtual Jewish History Tour · Jewish Virtual Library (American-Israeli Cooperative Enterprise)
  24. 24.South Africa (Jewish community profile) · World Jewish Congress
  25. 25.Plungyan: A Memoir (Jacob Yosef Bunka), Chapter III: The War and the Destruction of the Jewish Community of Plungyan · JewishGen Yizkor Book Project
  26. 26.Professor Michael Levitt FRS · The Royal Society
  27. 27.Michael Levitt, Robert W. and Vivian K. Cahill Professor of Cancer Research (Stanford Profiles) · Stanford University
  28. 28.Zinbryta: Article-20 procedure (EMA review confirms medicine's risks outweigh its benefits) · European Medicines Agency, 2018
  29. 29.Michael Levitt · Jewish Virtual Library (American-Israeli Cooperative Enterprise)

Fact-checked on September 24, 2026 by a separate AI fact-checking pass that re-opened the sources, with 8 corrections made. How we check

Suggest a correction