FEATURE
Tags: 

The life puzzle solver

Pavel Pevzner, chair professor of computer science, University of California, San Diego

Like many mathematicians Pavel Pevzner likes to solve puzzles. The puzzles he has chosen to solve are those of the biological world concerning the very basic stuff of life itself; proteins and DNA. Working away behind the old Iron Curtain he found out, before many people in the more open Western World, that biology presented some quite interesting and deserving puzzles. Since moving to the US he has become one of the pioneers of bioinformatics. In his case this has involved the search for algorithms that will unravel the mysteries of how cells work in terms of information processing rather than chemical factories.

 

His work at the University of California, San Diego crosses many disciplinary boundaries and has produced stunning work that has advanced our understanding of how processes like evolution or diseases like cancer actually work.

 

Professor Larry Smarr, a colleague at the University of California, San Diego, and director of the California Institute for Telecommunications and Information Technology (CalIT2) was just setting up the institute when Pevzner joined, and was delighted to have someone who could work in the kind of cross-disciplinary team that the organisation was designed to support. He says: ‘Pavel has taken the lead in creating, within CalIT2, a new centre for algorithmic and systems biology, bringing together many of the best people in UCSD in applying new approaches and algorithms to genomics, proteomics, cell signalling and so on. 

 

‘A lot of bioinformatics is quite pedestrian; how to use databases and things like that.  Pavel is one of the real scientific pioneers in what is emerging as a new field of science, which is living systems as information systems. The central dogma of DNA and RNA emerged as a way of writing down the winners of the genomic sequence that was appropriate to a particular eco-niche and transmitting that on to future generations. So evolution is really is central to the information systems. One of the things that Pavel has done, which is only beginning to be appreciated, is that you have to look at the entire history of the genome in order to understand the species.

 

‘He has a deep sense of how central evolution is to all things biological. His work on chicken, mouse and rat genomes have been covered articles in Nature; they are historical landmarks. He takes that understanding of evolution to look at how the DNA itself evolves. He has now started looking at cancer and found that cancer cells use the same techniques, but on a much shorter timescale. He is pulling out some very deep abstractions about how living systems work.

 

‘He is a delightful person. He has great humour and is always smiling. He is a very intense worker. He is a very private person. He is not out selling himself, rather he is very calm and competent; a “still waters run deep” kind of person. You have to discover he’s there and judge the depth of his insight. He does not have the outward brashness that a lot of Americans have.

 

‘All of our faculty are smart, but there are only certain people who you know are going to become leaders in science, and that is the sense I get about him.’

 


Pavel Pevzner

Pevzner was born in Kursk, a city about 300km south of Moscow. His mother was a teacher and his father was an electronic engineer. When he was a baby, his family moved to the town of Murom, about the same distance east of Moscow, which was the centre of the Russian electronics industry.

 

He had a slow start to his education. He says: ‘I was a very bad student until I was about 10 years old; I was only interested I having fun. I was a C grade average student. But then I started taking an interest in books.

 

‘The education system in the Soviet Union at that time was excellent, mainly because intelligent women did not have many options for careers, except as teachers or doctors, so there were a lot of excellent women teachers.’

 

When he was 14 he went to a special boarding school in Moscow, which was for children who are gifted in mathematics and physics. The school was founded by Andrey Kolmogorov, one of the leading mathematicians in the world on probability theory and topology. The school was based at Moscow University and Kolmogorov used to teach in the school three times a day.

 

‘I was very interested in mathematics, physics and biology. I don’t think you could say that I was a protégé - that is someone who is out of this world. I was very much of this world. 

 

‘It was very difficult at that time to get into a top university, because I was an ethnic Jew. In the mid 1970s it was nearly impossible for a Jew to get to Moscow University. There was some kind of unwritten rule that the percentage of students from every nationality admitted had to match the percentage of that nationality in the population, and Jews were typically over-represented among people applying. I had no intention of leaving the country when I graduated; I was quite comfortable there, but there was probably an impression among Russians that Jewish people wanted to leave.’

 

He failed to get into Moscow University despite having competed in the international mathematics Olympics. Instead he applied to the Russian Institute of Railway Engineers. This had an excellent applied mathematics programme and was full of very bright Jewish students who could not get into Moscow University. Pevzner had gotten used to a very demanding course at the Kolmogorov High School, so he found the undergraduate course quite within his capability. He did well and had already started doing his own research before he graduated. He even had some papers published on the optimisation of transportation systems.

 

He suddenly discovered that there were many more interesting problems in biology. He started looking for a postgraduate research opening in this area.

 

He says: ‘The system in Russia was a bit chaotic. There was no publication where you could see positions advertised; you simply had to get to know people and find out about things by word of mouth. I was very fortunately to find some people at a genetics institute. I went to talk to them for about an hour and I found what they were doing fascinating. At that time research was more of less private enterprise; you set up your research and then that became your PhD topic.’

 

Pevzner was eventually granted a PhD from the Moscow Institute of Physics and Technology in 1988. This was a very interesting time for Russia, which was in the throes of ‘Perestroika’ and ‘Glasnost’. Previously it was almost unheard of for Russians to be able to study or even travel abroad, particularly to the US. But one day a letter arrived from the Ministry of Education telling all Russian academics to consider spending some time abroad and that the government would pay for this. This came as a something of a surprise.

 

Pevzner says: ‘Many unusual things were happening at that time. There was no mechanism for a Russian to travel abroad and present something, let alone study. We were asked what university we would like to go to. I knew from the literature in the field of bioinformatics that if I was given the choice and the Russian government would pay and I just had to board the plane, then the University of Southern California was the place that I wanted to go. I put on the form that I wanted to work with Mike Waterman at USC.’

 

Pevzner did not take the chance that simply filling in the form would work. He started writing to Waterman. He says: ‘I told him that I was very interested in his published work and he sent me some new papers. In one of these, he had an open problem that I happened to solve. I sent him a solution and we started to communicate. He told me later that he could not believe that someone had solved this problem. Later he invited me to a conference and eventually he invited me to come to USC.

 

‘Of course the government never paid a dime towards it. It was quite funny that they should even make such an offer with a straight face.’

 

Pevzner spent two years at USC working with Waterman on his pioneering ideas on bioinformatics, which was an extremely small field at the time. They co-authored several important papers. At the time, Pevzner was working on DNA sequencing by hybridisation. He had read a report of a conference in a magazine about a crazy idea that some scientists had proposed on a quick way of sequencing DNA. It was rejected by many people, because it was thought of as impractical. Pevzner decided to ignore the cost of doing the experiment and just thought about how the data might be assembled.

 

He says: ‘I didn’t worry about the cost of implementing it I just thought about how you would decode the result of the experiment. The experiment is supposed to sequence DNA in a very indirect way and you have to assemble the result like a puzzle. I was always interested in how to solve puzzles and it turned out produce a perfect biological puzzle. I worked out how to solve this puzzle and published it in a US magazine. As a result of this, I was invited to a conference in Belgrade and I met the three scientists who had proposed the method. They included Ed Southern from Oxford University. I was surprised at how quickly this technology actually turned into a multi-billion dollar industry, which is DNA chips or expression arrays.’

 

After two years as a post-doc at USC, he was offered the chance to join the faculty. But at the same time he was offered a position as Associate Professor at Penn State University, and he decided this would be a chance to set up his own independent research group. After three years at Penn State, USC decided to make him an offer he could not refuse. He returned as a full professor, taking one student and another former student as a post-doc. After five years back at USC he moved again to his present position at San Diego.

 

He says: ‘I like to change places from time to time. San Diego was such a beautiful place and there was an opportunity to develop a new direction in the department, so it was a very attractive offer. There really is something unique in San Diego. It has Salk, Scripps and so on. There are many other cities with lots of institutions, but all these places are on the same street and are almost within walking distance of each other.

 

‘The place is very beautiful and I try to spend a couple of hours on the beach every day. It is warm, but if you are missing snow then its only a couple of hours to the mountains.’

 

Pevzner was now a professor of computer science as well as mathematics. He says: ‘The things I do is closer to computer science than mathematics. The distinction is not always clear, but I feel more comfortable in computer science. But I have a joint appointment and there is not such a distinction between the two. But in computer science departments, there is more of an emphasis on solving real problems than there is in mathematics departments, but the culture varies. Bioinformatics is really interdisciplinary. When appointments are made they are forced into one department, when in fact they should have a foot in several departments.’

 

Pevzner’s research in recent years has been around the evolutionary changes that take place in the genome. Some changes are through mutation, and he characterises these as being like the wind and rain weathering of a continent. There are other changes that are more dramatic, where the chromosomes rearrange themselves and these events are more comparable to earthquakes

 

He says: ‘There are definitively earthquakes, dramatic changes from time to time. The question arises as to where these changes fall and whether they are relevant to human disease. There is a theory called the “random breakage model” that states that all rearrangement is random and there are no hotspots. We have proved that this is not true and there are rearrangement hotspots. We have been trying to identify these and understand their relevance to human disease.’

 

Another important strand of his research has been in proteomics. As well as the DNA, there are many thousands of proteins in the cell, which are the workhorses of the cell.

 

He says: ‘DNA is seen as strings with a four letter alphabet, while proteins are described by a 20-letter alphabet (amino acids). But that is not quite true. Firstly there are more than 20 letters and also there are a large number of post-translational modifications that turn the proteins into different types of chemical entities. There is a whole world of post-translational modifications and so it is a gross oversimplification to see proteins as simply a 20-letter alphabet. It has to be viewed in terms of the alphabet of 20 amino acids and hundreds of modifications. This world of modification is still hidden from us. It is very important, for example how is blood type defined by this modification. Viral infection usually starts by recognising certain modifications on the cell surface. How these modifications are formed is a very complex and unsolved problem. We have found that this world of modification is much larger than previously imagined.’

 

Much of his work in this area revolves around finding algorithms to analyse the data from mass spectrometry. Huge amounts of data are produced, but it is difficult to see distortions in the data. Pevzner sees this as a puzzle, and he likes solving puzzles. He compares it to the children’s puzzle called ‘Where’s Waldo’. The information you want is in there all over the place, but you need to find a way of looking at the data that reveals those modifications all over the place.

 

He says: ‘All of a sudde,n if you have the right angle you see Waldo everywhere. We are looking for algorithmic solutions to this puzzle.’

 

Pevzner has seen a lot of changes in his career; as well as the political changes in his home country he has also seen his small field of bioinformatics explode onto the world’s scientific scene. He travels a lot now, but this has not made him keen to move to another university any time soon. It suits him to sit on the beach in San Diego, where he can spend some quiet time contemplating the solutions to the puzzles with which biology has presented him.

Pavel Pevzner CV

Education

 

1988 Ph.D. Moscow Institute of Physics and Technology Mathematics and Physics

 

Professional Experience

 

1986-1988 Ph.D. Trainee, Institute of Physics and Technology, Moscow, Russia

 

1985-1990 Scientist, Laboratory of Mathematical Methods, National Center for Biotechnology NIIGENETIKA, Moscow, Russia (formerly Institute of Genetics of Microorganisms VNIIGENETIKA)

 

1990-1992 Postdoctoral Research Associate and Lecturer

Department of Mathematics, University of Southern California,

 

1992-1995 Associate Professor, Department of Computer Science,

The Pennsylvania State University, PA

 

1995- 2000 Professor, Departments of Mathematics, Computer Science, and Molecular Biology

University of Southern California, CA

 

2000-present Ronald R. Taylor Chair Professor of Computer Science

University of California, San Diego, CA

 

2002-present Adjunct Professor of Mathematics

University of California, San Diego, CA

Analysis and opinion
Feature

Robert Roe looks at research from the University of Alaska that is using HPC to change the way we look at the movement of ice sheets

Feature

Robert Roe talks to cooling experts to find out what innovation lies ahead for HPC users

Analysis and opinion