Thursday 18 July 2019

Dogged does it.

As I push on towards retirement I am decluttering in a piecemeal kind of fashion. Off-loading stuff that I cannot imagine anyone having a use for.  My track record is not good for imagining uses for data. I was only annoyed when Craig Venter published a clatter of ESTs from a human brain: because ESTs are partial and of crappy quality and all I wanted were full length human genes which had been accurately sequenced. Wronnnng! ESTs are dead easy to generate and, because of their monstrous regiments, can be used to answer all kinds of interesting questions including one of my own [PMID: 12885301]. I can do methodical but I'm better at butterfly. The field I chose to work in 30 years ago doesn't give credit for drudges - folks who can plod through hundreds of samples and maintain the same standards of accuracy and reproducibility from start to finish. In bioinformatics, you get credit for scaling up an analysis by writing software that can do the heavy lifting, scouring and sorting through scads of data.

There is another class of scientists which I have characterised as A Good Pair of Hands. Like my student Grace who was the only one in the Microbiology class who could do a smart, reliable diagnosis with Gram stain. Turned out that she had spent the previous summer in an equine pathology lab doing dozens and dozens of Gram stains. I love doing Gram stains, she said, as she ploughed through a stack of them. As with so much in life practice makes perfect, but I knew that her managers in the equine lab developed an absolute trust on her reliability, accuracy and efficiency.

My brother tells a story about quality control QC technicians in a Volvo factory. The quality of the assemblies was so good that the QC checkers were falling asleep, day-dreaming and thinking about the weekend's soccer. Th QC manager decided to introduce a bunch of deliberate errors in the pipe-line: they came often enough and random enough to keep the QC techs on their toes and able to catch, not only the known errors, but also the unknown / real errors on the system.  I was reminded of this because of a recent molecular evolution report out of Tucson, AZ.

Michael Worobey, the PI, heard about an archive of 50+ year old paraffin-embedded tissue samples stored in cardboard boxes in the University of Kinshasa in the DRC / Democratic Republic of Congo [previously Zaire; prevprev Belgian Congo). Let us give thanks to inertia and a tolerance of skuzz and clutter that these slides hadn't long ago been tidied away to whatever serves for landfill in central Africa. Like all the unique irreplaceable photographs of olde time Tyneside that my pal Roy was unable to save from discard and destruction. Worobey has a track record in medical molecular archaeology, extracting DNA from old biopsies with painstaking care, so that the threadlike fragile genetic signal can be filtered out of the noise of time-degraded and potentially contaminated biological material. The samples were probably preserved by histologists who wanted to see what they looked like down a microscope. Nobody then knew that the DNA was useful / informative let alone how best to preserve that aspect of the tissue.  But there was a helluva lot of samples: N = 1652 to be precise, and someone had to process them all. That task fell largely to Worobey's Good Pair of Hands a fellow called Thomas D Watts. They were hoping for some ancient HIV on the assumption that these central African samples would predate the HIV/AIDS epidemic that started scything through young gay men in New York and California in the early 1980s.

Well they ran through all the boxes and boxes of material and came up with one (1) positive for HIV - a 38 y.o. male who presented to the hospital in 1966, where a peanut-sized chunk of a lymph node was taken for further study. Despite the lottery of history [116/123 of the plays of Sophocles have been lost] and because they have really good techniques for assembling fragments of DNA gathered in adverse circumstances, Team Tucson were able to obtain a more-or-less complete HIV genome sequence from this sample and compare it to all the subsequently evolved HIVs. There are, as yet, no known older HIV sequences which are mostly complete; DRC66 is the oldest. Evidence, from phylogenetic trees and molecular clocks, is pointing to the start of the 20thC for The Event that brought HIV-I across the species divide from chimpanzees Pan troglodytes to us Homo sapiens. Watch this space as researchers get dustier colonial times boxes from peculiar places to find older HIV positive sequences. These will almost certainly be confirmatory, it would be surprising if something startling emerges from similar studies. But the Tucson study is valuable as an anchor-point for the inferences of phylogenetic trees generated by computer. If this is HIV-I, there is an HIV-2, right? Indeed so, HIV-2 crossed in humans from the sooty mangabey Cercocebus atys. That strain is much less virulent and infective, so is rarely seen outside of West Africa.

Remember Ebola? That had us all up in a flap 2013-2016 and was almost a case of too little, too late? The mainstream media are tired of Ebola and don't want to be seen to cry wolf. Nevertheless, Ebola is back and has been flagged with global emergency status by the World Health Organisation. And the epicentre of the latest holocaust? DRC!

No comments:

Post a Comment