I have been working in the areas of bioinformatics and molecular evolution since 1985. My main work has been on methods and software for DNA and protein sequence alignment. This grew from a need in the late 1980s to make multiple alignments of sets of related sequences, in order to carry out further analyses such as phylogenetic reconstruction. Up to about 1987, if you wanted to take a set of sequences and make a multiple alignment, you had to do this using sheets of paper and coloured pens, or, if you were well off, a word-processing program and a computer. This was tedious and error-prone, so there was a clear need to automate this process. It turned out to be surprisingly difficult to write computer programs to do this, especially considering that humans could do it easily, even if painfully slowly.

My first work in this area was when I was a post-doc in Paul Sharp’s laboratory in Trinity College, Dublin, Ireland. I wrote a series of programs, called Clustal, that could carry out reasonably accurate multiple alignments but could do this quickly and simply, even on an old IBM PC with an 8088 processor and very little memory. This program quickly grew in popularity and has evolved through a long series of jumps and steps to become the world-wide standard software for this analysis. Because we started it off on such a miserably weak computer, as computers became very powerful in the 1990s, this more than compensated for the great explosion in the numbers and lengths of sequences that most people needed to align. The first papers describing these programs were published in 1988 and 1989. Later, I moved to the EMBL in Heidelberg, Germany, where I published a new, improved version of the program with my colleagues Alan Bleasby and Rainer Fuchs. This was ClustalV (the V stood for 5 as this replaced the earlier programs which were called Clustal1 to Clustal4) and it was available for IBM PC, Mac, Unix, and VAX/VMS computers.

In 1994, with my colleagues Julie Thompson (now in Strasbourg, France) and Toby Gibson (still at EMBL, Heidelberg), we produced ClustalW which is now the main software for carrying out multiple alignments along with ClustalX, which is essentially ClustalW with a nice graphical interface and many extra visual features. ClustalX was also made in collaboration with Francois Jeanmougin and Frederic Plewniak, also in Strasbourg. These programs are used hundreds of times every day to produce multiple alignments, and the papers that describe them get cited about 50 times a week. There are dozens of WWW servers that use these programs and they are resold in many commercial packages.

More recently, I have been working on a new method, called T-Coffee which was invented by a Ph.D. student in my group at the EMBL, Cedric Notredame. T-Coffee is slower than ClustalW, but it produces more accurate alignments and allows us to mix different types of data together such as protein structures, ESTs etc. I now teach biochemistry in University College Cork, Ireland, and my group works on not only multiple alignments but also on bacterial genomics and the application of multivariate analysis to analysing microarray data.

Des Higgins, Ph.D.
Department of Biochemistry
University College Cork
Cork, Ireland