Skip to main content
Premium Trial:

Request an Annual Quote

Tandy Warnow, PI


For ‘wild woman’ Tandy Warnow, evolution is a whodunit. The comp sci gumshoe has set her sights on detailing how early genomes begat the genomes you know and love today.


By John S. Carroll


Tandy Warnow loves to dance. “It’s my new passion,” she declares excitedly, eyes flashing as she rattles off her favorites: swing, Zydeco, salsa, lindy.

For Warnow, co-director of the Center for Computational Biology and Bioinformatics at the University of Texas, Austin, dance is an extemporaneous interplay between two people of complementary strengths; an ongoing discovery of movement. “It’s what I can do with someone else,” she says, “that makes it so wonderful.”

It’s also a metaphor for her work in computational phylogenetics. Warnow, 47, applies her expertise in math and computer sciences with a tightly knit group of collaborators — sleuths of genomic data who exchange ideas on their quest to reconstruct complex evolutionary histories of organisms.

It’s a tiny band of explorers. Warnow names biologists Robert Jansen and Randy Linder at UTA and Bernard Moret and David Bader at the University of New Mexico as the closest members of her inner circle. The group gelled a little more than three years ago, shortly after Warnow left her tenured position at the University of Pennsylvania for her current UTA post.

Together, they’ve embarked on the ambitious project of large-scale phylogenetic analysis, reconstructing evolutionary history using the flood of new molecular data. One of their main projects is tracing phylogenies for whole genomes. Optimization problems for whole-genome phylogenetics are enormously hard — much harder than the corresponding problems for DNA sequence phylogeny.

For this problem, the team has developed software that is 5 billion times faster than the best software alternatives now available. That’s good, Moret says, but it still falls far short of where they need to be to efficiently reconstruct infinitely complex phylogenies.

To understand how organisms evolved, their work looks at how whole genomes evolve. And to see how a gene evolved they study how DNA sequences developed, theoretically reconstructing the mutations that occurred over time.

Those orderly branches in most evolutionary tree models just don’t work in biology, says Warnow, whose algorithms are needed to take this messy, massively complex data and make sense out of it. “It’s amazing,” she says, “that we can do it at all.”

A Career Evolves

Ultimately, their work fits into an enormously ambitious project called the Tree of Life at the University of Arizona, which eventually hopes to understand how all living organisms relate to each other.

Warnow also works directly with Lisa Vawter, a biologist doing drug design work at Aventis Pharmaceuticals. You can include her in Warnow’s list of enthusiastic dance partners.

“She’s a wild woman,” says Vawter with a laugh. “Tandy has more energy than anybody else I know.” And she brings that trademark electricity to everything she does — whether it’s cooking up a new algorithm, singing arias from Die Zauberflöte (she was a trained coloratura soprano in her 20s), or “spouting proofs” as she hikes an Appalachian trail.

Warnow’s penchant for broad intellectual inquiry keeps her hopping. In two intense months last summer, the divorced mother of two daughters crisscrossed Europe to attend a trio of scientific conferences, jumped to California to go dancing, and juggled teaching and research duties in Austin. There’s also a boyfriend to keep track of: Jon Bentley of Avaya Research Labs.

Through it all, she keeps pushing the envelope with her partners on phylogenesis.

Personal relationships have always been important in Warnow’s life. She first got involved in computational biology when she was completing her doctorate in mathematics at the University of California, Berkeley, largely because of the influence of the late Gene Lawler, who acted as her advisor. Ironically, her twin sister, Kimmen Sjölander, also landed in bioinformatics, and now teaches at Berkeley.

Perhaps there’s an algorithm that can help explain the trajectory of Warnow’s own life and career based on the evolution of her family.

Her mother, Joan Blewett, was an archivist of 20th century physics, and her stepfather was a noted accelerator physicist. But Warnow might identify more closely with her grandfather, Mark Warnow, the bandleader for the Lucky Strike Hit Parade Radio Show, or her great-uncle, composer, musician, and inventor Raymond Scott, who shares a link on her website.

Warnow’s fascination with evolution led her to explore how languages have developed. Six years ago, she used her expertise in designing algorithms to model how European languages may have branched out of a common Indo-European tongue spoken 5,000 years ago. Not only did the study offer suggestions for the way languages like English and French developed, it also proposed links between tribal migrations and the ways languages influenced each other along the way.

“She’s intellectually very curious,” says UTA’s Randy Linder, who connected with Warnow shortly after she moved from Penn. “She has a sharp mind and she lets nothing slip by.”

In much the same way, Warnow’s computational phylogeny group provides models for the way taxa developed and influenced each other as they evolved. But the amount of data that has to be analyzed is so massive, it can take months for a program to run. Between Warnow’s algorithms and her colleagues’ input on biology, software, and hardware, they hope to reduce that time to days, hours, and ultimately minutes or seconds.

“Understanding evolution is the key to understanding how to treat diseases, climate changes, plants and crops, and all sorts of things like this,” says New Mexico’s Moret. “It’s a tool through which evolutionary biologists can put their knowledge into an organized whole and gain an understanding of the big picture.”

Bonding Experience

Warnow doesn’t just seek out collaboration, she’s constantly breaking down the walls that isolate scientists. That’s one reason that she is so enthusiastic about UTA. Not only is the computer science department strong, “but what I didn’t know was how remarkably congenial it was.”

She places a big emphasis on congeniality, finding respectful ways to disagree and establish common ground for a department filled with colorful individuals, including some of her best friends.

Case in point: The phylogenetics group is sticking with open-source software, says Warnow, to help any other scientist in the field take what they’re doing and improve on it. The more minds that work on the problems and limitations, she says, the faster they will all find solutions.

And she’s already very fast.

“After you ask her a question, you get an answer, and it’s the right answer,” says Moret. “Some others take a week, she takes seconds.”

Moret has been working with Warnow, off and on, for the past 12 years. And over the past three years of close collaboration and shared grants, he says, they manage to interact effectively and frequently, closing the distance between the two universities with late-night telephone conversations and frequent e-mails.

By this time, the scientific dance between Warnow and Moret can be played out over long distances. And the interplay of computer code and intellectual quest will go on for years to come, with Warnow weaving her work into a scientific salsa of discovery.

The Scan

Study Links Genetic Risk for ADHD With Alzheimer's Disease

A higher polygenic risk score for attention-deficit/hyperactivity disorder is also linked to cognitive decline and Alzheimer's disease, a new study in Molecular Psychiatry finds.

Study Offers Insights Into Role of Structural Variants in Cancer

A new study in Nature using cell lines shows that structural variants can enable oncogene activation.

Computer Model Uses Genetics, Health Data to Predict Mental Disorders

A new model in JAMA Psychiatry finds combining genetic and health record data can predict a mental disorder diagnosis before one is made clinically.

Study Tracks Off-Target Gene Edits Linked to Epigenetic Features

Using machine learning, researchers characterize in BMC Genomics the potential off-target effects of 19 computed or experimentally determined epigenetic features during CRISPR-Cas9 editing.