Andrew McLachlan, former Group Leader in the LMB’s Structural Studies Division, died on Thursday 7th July 2022, aged 87. Andrew was a pioneer of structural bioinformatics and in interpreting protein structure on the basis of sequence alone. He developed several original computational methods for analysing and comparing sequences.
Andrew was born on 25th January 1935 in London. He grew up in Hampshire, near Winchester and from an early age was interested in science, particularly astronomy. He was educated at Pilgrims’ School in Winchester, the cathedral choir school, as a non-singer, and in 1948 he gained a scholarship to Winchester College, specialising in natural science, mostly physics and mathematics. In 1953, he was awarded a scholarship to Trinity College, Cambridge to read mathematics and theoretical physics, and graduated in 1956.
Andrew then started research with Christopher Longuet-Higgins in the Department of Theoretical Chemistry, University of Cambridge. There he worked on the molecular orbital theory of aromatic hydrocarbon radicals, their electron spin distributions and their electron spin resonance spectra. He was awarded his PhD in 1959. Following this, and with a Harkness Fellowship from the Commonwealth Fund of New York, he was able to work and travel in the United States for two years. He spent the first year with Harden McConnell, a leading figure in electron spin resonance, at the California Institute of Technology (Caltech), and the second year at the Chemistry Department of Harvard University with E. Bright Wilson, an outstanding theoretical chemist with interests in infra-red spectra.
Andrew returned to Cambridge in 1961, to continue his work on theoretical chemistry, and he became a College Lecturer in physics at Trinity College, a position he held until 1986. In 1964, he was appointed Assistant Director of Research in the Chemistry Department and in 1965, University Lecturer in Organic Chemistry. By this time, though, he had already agreed in principle with Max Perutz to join the LMB. From the early 1960s, Andrew had become increasingly interested in molecular biology, and had many discussions with Francis Crick in 1965, when both Francis and Andrew were visiting Caltech. In the summer of 1965, Andrew attended the Cold Spring Harbor course on bacteriorphage genetics. He also learned more about molecular biology from his friend at Trinity College, Richard Watts-Tobin, who was working with Sydney Brenner and Leslie Barnett at the LMB on the genetic code.
In 1967, Andrew joined the LMB as a senior scientist. He initially worked with Max’s group on the mechanisms of haem-haem interaction in haemoglobin, and his first task was to build X-ray map models of haemoglobin, working with Judd Fermi. However, Andrew’s interests soon turned to more general problems of protein structure and in 1972 he began a series of studies on the amino acid sequence of proteins, and later extended this into studies of DNA sequences. He developed several original computational methods for analysing and comparing sequences and many computer systems in the early 1980s were based on these procedures. In particular, he produced methods for the rigorous statistical analysis of correspondence between sequences which gave a unique precision to his work. One of his most important contributions was the idea that sequences of proteins evolved by gene duplication and adaption of an ancestral motif and that this can be detected in present day proteins by extracting repeated regularities. Andrew’s structural analysis of coiled-coil proteins, with Murray Stewart, led to the discovery of 14 proposed actin-binding sites in tropomyosin and the influential “rolling” model for the regulation of muscle by tropomyosin. In the mid 1980s he made a careful analysis of the amino acid sequence of nematode myosin, derived from the nucleic acid sequence determined by Jon Karn. This revealed a remarkable pattern of multiple repeats and alternating charge clusters which he interpreted in terms of a staggered linear model for the thick filament of muscle. In the late 1990s, Andrew developed a new family of mathematical methods and computer programs for solving or refining molecular structures from X-ray diffraction data.
Andrew mainly worked on his own, without a significant group, but he was a key collaborator with many other scientists, particularly at the LMB, including: with Alan Weeds and Murray Stewart on tropomyosin and troponin; with John Walker on serum albumin; with Richard Henderson on the helical arrangement of bacteriorhodopsin; with Fred Sanger and Rodger Staden on codon usage; with Aaron Klug on zinc fingers; and with David Eisenberg on sequence profiles. Andrew’s research encompassed mathematics, design and testing of computer programs, and solutions of actual molecules. In 1989, Aaron Klug wrote: “Although there are many people who work in the field of protein sequence analysis, most of the work is essentially descriptive. The difference in McLachlan’s approach is that the methods he has developed are powerful, practical, and at the same time statistically rigorous. I think it is true to say he holds a unique position in the field” (Correspondence from Aaron Klug to the MRC, 21 September 1989. Biographical File, LMB Archive).
In October 2000, Andrew became an Emeritus Scientist, and finally retired from the LMB in 2006. He was elected a Fellow of the Royal Society in 1989, for his work on the analysis of protein and nucleic acid sequences and of three-dimensional protein structures.
Alan Weeds, a former LMB Group Leader and scientific collaborator of Andrew’s, commented: “Although Andrew was not a conventional group leader who had a team working with him, his strength was in his many successful collaborations in the lab and elsewhere and the invaluable assistance he gave to many LMB scientists.” He also added: “He was greatly respected as a wise Fellow of Trinity College and lunched virtually every day following his wife’s sad death a few years ago.”
Repeating sequences and gene duplication in proteins. A. D. McLachlan. Journal of Molecular Biology 64: 417-437, 1972.
Fourteen actin-binding sites on tropomyosin? M. Stewart, A. D. McLachlan. Nature 257: 331-333, 1975.
Periodic features in the amino acid sequence of nematode myosin rod. A. D. McLachlan, J. Karn. Journal of Molecular Biology 164: 605-626, 1983.
Profile analysis: detection of distantly related proteins. M Gribosky, A. D. McLachlan, D. Eisenberg. PNAS 84: 4355-4358, 1987.