Kevin MacDonald, Ph.D.

CSULB, Department of Psychology

Human Biodiversity

The following appeared on Human Biodiversity, an email discussion group:

Regarding the NYT story on the report by Genaissance.com documenting big individual and found in genome studies that actually look at more people than Celeron's celebrated work. Might this provide a biological basis for race?

Response by Peter Frost:

I wouldn't get my hopes up. Human and ape SNPs have already been compared
and there is so much species overlap in their variability that it's
impossible to tell individual humans and apes apart.

FROM Vincent Sarich

TRYING TO STRAIGHTEN THINGS OUT:

THIS JUST ISN'T TRUE. APE (CHIMP) DNAs DIFFER FROM OURS BY ABOUT 1.3% IN SEQUENCE. THAT IS, IF YOU LINE UP THE DNAs, YOU WILL SEE A DIFFERENT NUCLEOTIDE (A, G, C, T) IN THE HUMAN AND CHIMP VERSIONS AT ABOUT ONE POSITION IN 70. IF YOU DO THIS WITH TWO HUMANS, THE FIGURE IS ABOUT ONE IN A 1,000. NO WAY ARE YOU GOING TO HAVE TROUBLE TELLING HUMANS AND CHIMPS APART. AND, IF YOU WENT BACK TO THE LOW TECH WORLD I INHABITED 25 YEARS AGO, I COULD DEMONSTRATE THIS FOR YOU IN AN AFTERNOON. I WOULD SIMPLY TAKE, SAY, 10 SERUM SAMPLES (5 HUMAN, 5 CHIMP), PUT THEM IN SLOTS IN A POLYACRYLAMIDE GEL, AND PUT THE GEL IN AN ELECTRIC FIELD. THE MANY PROTEINS IN EACH SAMPLE WOULD THEN MIGRATE AT DIFFERENT RATES DEPENDING ON THEIR SIZE AND CHARGE, AND AFTER A FEW HOURS I WOULD TAKE THE GEL AND STAIN IT. I COULD THEN SEE BANDS (ABOUT 25), EACH REPRESENTING A DIFFERENT PROTEIN, AND THE LARGE MAJORITY (ABOUT 20) OF THOSE WOULD SHOW DIFFERING MOBILITIES IN THEIR HUMAN AND CHIMP VERSIONS -- IN OTHER WORDS, DURING THE 5 MILLION YEARS OR SO THAT THE HUMAN AND CHIMP LINEAGES HAVE BEEN SEPARATED FROM ONE ANOTHER, ONE OR THE OTHER, OR BOTH, WILL HAVE ACCUMULATED MUTATIONS (HERE AMINO ACID SUBSTITUTIONS) CONFERRING DIFFERING CHARGES AND/OR SHAPES TO THEIR PROTEINS. IT WOULD BE IMPOSSIBLE TO MISTAKE A HUMAN SAMPLE FOR A CHIMP SAMPLE, OR VICE VERSA. NO HUMAN WOULD DIFFER FROM ANY OTHER HUMAN BY AS MUCH IT DIFFERED FROM ANY OF THE CHIMPS, AND VICE VERSA. AND IT WOULDN'T BE ANYTHING APPROXIMATING A CLOSE CALL.

This is incidentally the same problem we get with variability in a lot of other junk DNA. There's no natural selection acting on these genes to keep different human populations (or even primate species) distinct. Since only a limited number of variants are possible (5 in the case of SNPs), the same variant will arise independently in unrelated populations.

FIRST, A TRIVIAL POINT: IT'S 3, NOT 5 (IF IT'S AN A, IT CAN GO TO G, C, or T).IT IS OF COURSE TRUE THAT THE SAME VARIANT CAN ARISE IN UNRELATED POPULATIONS, BUT THE LIKELIHOOD OF IT ACTUALLY DOING SO IS LOW -- NOT ZERO, OF COURSE, BECAUSE EVEN DNA VARIABILITY IS LIMITED, BUT LOW. SPECIFICALLY, TAKE OUR CHIMP/HUMAN EXAMPLE. IF TODAY THEY DIFFER AT ONE POSITION IN 70, THEN THE NEXT MUTATION IN EITHER LINEAGE HAS, SIMPLIFYING SLIGHTLY, ABOUT ONE CHANCE IN 210 (70*3) OF DUPLICATING ONE THAT IS ALREADY PRESENT IN THE OTHER LINEAGE.

I predict that we will find much more SNP variability within races than
between races. And this finding will be one more nail in the coffin of the
race concept.

THIS IS A VERY SAFE PREDICTION, AS WE ALREADY KNOW (AND HAVE KNOWN FOR A LONG TIME) THAT WITHIN-GROUP VARIABILITY IS MUCH LARGER THAN BETWEEN GROUP VARIABILITY -- ALTHOUGH AGAIN I THINK THAT THE WAY WE TEND TO SPEAK ABOUT IT CONFUSES THE NON-GENETICISTS -- IF SOME OF THE COMMENTS FROM VARIOUS MEMBERS OF THIS GROUP ARE REPRESENTATIVE. SAYING THAT W-G VARIABILITY IS GREATER THAN B-G VARIABILITY DOES NOT MEAN THAT TWO INDIVIDUALS IN ONE GROUP (POPULATION/RACE) WILL BE MORE DIFFERENT FROM ONE ANOTHER THAN EITHER IS FROM AN INDIVIDUAL IN ANOTHER GROUP (POPULATION/RACE). BUT IT, IN CONTRAST TO THE HUMAN/CHIMP COMPARISON, IS A CLOSE ENOUGH CALL THAT IT WILL OFTEN BE TRUE THAT THE NUMBER OF OBSERVED DNA DIFFERENCES BETWEEN TWO INDIVIDUALS OF POPULATION A WILL BE FEWER THAN BETWEEN EITHER OF THEM AND ONE FROM GROUP B. WHAT WE MEAN, AGAIN SIMPLIFYING SLIGHTLY, IS THAT, ON THE AVERAGE, TWO HUMANS FROM DIFFERENT POPULATIONS WILL DIFFER FROM ONE ANOTHER BY ABOUT 1 NUCLEOTIDE IN A 1000, WHILE TWO FROM THE SAME POPULATION WILL DIFFER, AGAIN, ON THE AVERAGE, BY 10-15% LESS -- SAY 1 IN 850 OR 900.

BUT ALL THIS HAS, BASICALLY, NOTHING TO DO WITH THE RACE CONCEPT. RACES ARE NOT SPECIES, THEY ARE YOUNG, AND GENE FLOW AMONG THEM HAS ALWAYS EXISTED. THERE ARE THUS VERY FEW VARIANTS IN OUR SPECIES WHERE ONE VERSION (ALLELE) IS PRESENT IN HIGH FREQUENCY IN ONE POPULATION, AND ANOTHER AT HIGH FREQUENCY IN ANOTHER POPULATION. "HIGH" HERE IS, SAY, 80% OR MORE. BUT WE ALSO KNOW THAT RACIAL VARIATION IS PRESENT, AS WE CAN, WITH APPRECIABLE ACCURACY, LOOK AT AN INDIVIDUAL AND RECOGNIZE WHERE HE [OR HIS RECENT ANCESTORS] RESIDE [OR RESIDED]. AND THIS IS EVEN TRUE FOR A SUBSTANTIAL PERCENTAGE OF "HYBRIDS". AND THE FACT THAT W-G DIFFERENCES CAN BE GREATER THAN B-G DIFFERENCES FOR MANY FEATURES DOES NOT PREVENT US FROM PICKING UP ON THOSE FOR WHICH IT IS LESS, OR NOT AT ALL, LIKELY, AND USING THEM TO DO THE EQUIVALENT OF OUR LOOK/SEE GROUPINGS USING VARIATION AT LEVELS "CLOSER TO THE GENE" -- FROM CRANIAL/FACIAL MEASUREMENTS TO BLOOD GROUPS TO SERUM PROTEINS TO, EVENTUALLY, SNPS.

WE CAN ALSO "PREDICT" WHAT PROPORTION OF OUR SNPS WILL TEND TO BE POPULATION SPECIFIC. THEY WILL BE THOSE THAT HAVE ARISEN SINCE THE OUT-OF-AFRICA EVENT SOME 50,000 YEARS AGO -- ASSUMING, OF COURSE THAT THERE WAS SUCH AN EVENT (IF THERE WASN'T, THIS SORT OF ANALYSIS WILL DEMONSTRATE THAT AS WELL). WE KNOW THE RATE AT WHICH NUCLEAR DNA MUTATIONS ACCUMULATE -- ABOUT 1% OF THE 3 BILLION BASE PAIRS PER 8 MILLION YEARS. OVER 50,000 YEARS THIS TRANSLATES TO 1 SUBSTITUTION PER 16,000 NUCLEOTIDES THAT WILL TEND TO BE POPULATION-SPECIFIC. THIS DOESN'T MEAN THAT POPULATION A WILL HAVE 100% CYTOSINE AT POSITION 2,345,788,666, WHILE ALL THE OTHERS HAVE 100% ADENINE -- BUT THAT IT WILL HAVE THE QUANTITATIVE EQUIVALENT SPREAD OUT OVER SEVERAL POSITIONS (THAT IS, AND VERY CRUDELY, POPULATION-SPECIFIC SUBSTITUTIONS AT 5 POSITIONS, EACH AT A FREQUENCY OF 20%), SUCH THAT ANY GIVEN INDIVIDUAL IN THAT POPULATION WILL HAVE ONE OR MORE OF THESE, AND THEREFORE BE IDENTIFIABLE AS A MEMBER OF THAT POPULATION. THE IDEA HERE IS CORRECT -- I'M NOT SO SURE ABOUT THE MATH -- OTHERS CAN CHECK ME OUT ON IT. BUT THE PRECISE NUMBERS HERE ARE NOT THE POINT, WHICH IS: THERE MUST BE, AND IS, POPULATION-SPECIFIC VARIATION PRESENT AT THE DNA LEVEL, AND WE CAN CALCULATE HOW MUCH THERE IS GOING TO BE.

ENOUGH FOR NOW:

VINCENT SARICH