Genes for the Next Generation

By Paul VanRaden

This follows previous reports on human genetics from Paul:

1983. Quotes from leading thinkers 1693-1983

1985. Proposal for human progeny testing

2023. The embryos that do not become babies

Topics:

1) Any genetic progress since 1985?

2) Mate choice and genetic variance

3) Genetic counseling

4) Genomics

5) Prediction

6) Advice

7) My choices and attitudes

8) References (general, companies, progress, historical)

Any progress since 1985?

After writing a report on human genetics in 1985 I expected that human DNA would not improve quickly, and it did not. Compared to the DNA that our parents’ and grandparents’ generations inherited from their ancestors, humans on average have made no genetic progress. Each generation of life has new mutations that mostly make DNA a little worse. Many embryos and some babies or children inherit mutated genes or defective chromosomes that cannot sustain life. Then, natural selection prevents them from becoming adults and giving their genes to the next generation. That same, slow process of natural selection took 6 million years starting with ape DNA to give us our modern human DNA. Gradual evolution is a natural but not a pleasant process for those whose genes cannot sustain life.

By contrast, the DNA of many animals and crops continues to improve rapidly. Instead of 25 years per generation, other species reproduce in fewer years or can even have multiple generations per year. Cows today produce twice the milk of cows 50 years ago. Chickens bred for egg production lay 79% more eggs by 60 weeks of age. Chickens bred for meat production grow 4 times faster and have >50% better feed efficiency. Corn fields yield 2.3 times as much corn, wheat fields 2.2 times the wheat, and rice fields yield twice the rice. Genetic selection accounts for most of the progress for each of those trends in productivity. That progress in agriculture helps easily feed 8 billion people on a planet that struggled to feed 4 billion people 50 years ago.

Humans, animals, and plants are all improving slowly compared to computers. Even with the same DNA, people now can get much more done than previous generations by using computers and other electronic devices redesigned to be twice as efficient every two years. But some people may get less done by playing video games or scrolling their phone screen all day. People who study hard and learn new skills for decades may soon be no match for faster computers with more memory that hold more information than you might read in a lifetime and be reprogrammed almost instantly to do other complex tasks. As the gap between their skills and ours increases, many future jobs for humans might be helping the computers and robots rather than them helping us.

People enjoy and expect to see progress in food, housing, transportation, medicine, and entertainment during their lifetimes. Because each generation uses their DNA for several decades, the next generation may wonder why this generation was not more careful. If human DNA got worse each generation such as from exposure to radiation or chemicals, society might see that as a major emergency that needs to be fixed. Similarly, society could see that improving DNA is a major opportunity. One reason that fertility clinics have more business is that male fertility is declining.

Children could have chromosomes that provide better health than the damaged chromosomes from previous generations that may have harmed their parents’ or ancestors’ lives. You could improve human DNA in the next generation by giving your children chromosomes free of obvious, known defects that may have harmed you or could harm them. Or you may simply want your children to inherit better chromosomes than your parents gave you.

Mate choice and genetic variance

Genetic variance can increase in a population if mates choose partners more like themselves or can decrease if opposites attract. Simple genetic theories often assume a randomly mating, very large, unselected population with very little inbreeding or changes in allele frequencies. Then the genetic mean and variance remain stable except for some mutation and some natural selection. In real populations, genetic means often change due to selection, migration, or genetic drift caused by inbreeding or sampling in small populations. Genetic variances also can change due to those same factors or from nonrandom mating.

Humans and many other species have a natural desire to combine their own genes with the fittest other genes available in the population. That strategy increases the chance that their progeny will survive and reproduce. Thus, people have a natural tendency to seek out the best mate or spouse they can find. Humans are generally a pair bonding species whereas in some other species one male wins a battle and then has several mates.

Many species have social hierarchies such as a pecking order that affects which individuals get first choice of food, shelter, or mates. With pair bonding, higher ranking males and females may pair more often and lower ranking males and females may pair more often than chance. For a human example, imagine the process of choosing a prom date. Those who want to go often want to take a date but who should that date be? Ideally prom dates like each other or have similar interests or attitudes. The two most likeable people might pair off and go with each other, leaving those of us that are less likeable to search for someone who likes us, wants to go, and might say yes. That process can cause positive assortative mating and positive correlations of social status score among prom dates, spouses, or mates.

Suppose that pairs choose each other by stature, with the shortest together and the tallest together. Genetic variance for stature would increase in the next generation if no child had one short and one tall parent. Heritability of stature is high, above 50%, but would also increase each generation due to more genetic variance but the same environmental or error variance. Each new generation would include more very short and more very tall people than the previous even if the mean did not change. If people would pair by income, then the rich would get richer and the poor would get poorer each generation if progeny inherit wealth via their genes or directly as cash from their parents."

In real life, people with more chances to meet also are more likely to pair. In the past, mates paired mostly with those who grew up nearby because that is who they met. Now, people travel more and can find each other using the internet, phone, or other communication. With more options and specialized jobs, mating may become even more assortative. Pairs might both be in medicine, or food service, or scientists, or musicians, or athletes, or movie stars. Most people are not surprised if children of athletes become athletes, children of movie stars become movie stars, the son of a President becomes a President, the first-born son of a King becomes a King, etc.

Much pairing might be on social status, resulting in children whose parents are both of higher status or both of lower status. The caste system in India previously required pairing within the same social class until banned in 1947. Nearly all pairs on earth today both grew up in the same nation. Similar mating systems create bigger gaps between the haves and the have nots than would normally occur. But deciding who to marry and deciding what genes your children should have could be 2 separate choices. That would allow parents of any status to find genes that your children might like.

Genetic counseling

To better understand genetics, heredity, and risks for their families, potential parents may get counseling from a trained specialist. Genetic defects occur frequently in humans and in all other species. Some individuals inherit simple genetic conditions caused mainly by a defect in one gene. Discovery of those conditions began after 1900 when Mendel’s 1860 paper on inheritance in peas was rediscovered.

Tests are now available for hundreds or thousands of genetic defects in humans. Most may have no effect on parents that carry only 1 defective copy of the gene but severely affect progeny who receive that same genetic defect from both parents. Those are known as recessive traits. In recent decades genetic counseling may be available to help couples better understand such conditions, decide on testing, avoid giving their children severe genetic defects, or discover if treatments are now available. Such defects in other species such as cattle often are quickly selected out of the population within a few generations because treatments may not be available or affordable.

Many embryos have abnormal numbers of chromosomes that are very harmful but are hard to predict in advance, except that older mothers have more risk. Humans should have 23 pairs of chromosomes, with one of each pair from the mother and one from the father. When fertilization occurs, the egg or the sperm often are missing one of the 23 chromosomes or may carry an extra copy of a chromosome because the earlier biological process of creating each egg and each sperm has a lot of errors.

Within your cells, the two sets of 23 chromosomes that you received separately from your mother and father line up in pairs before forming each gamete cell (sperm or egg). One of the two chromosomes in each pair is supposed to be randomly selected, but some chromosome pairs stick together: both go to one gamete and none to the other. In some cases, a chromosome from either parent may contain an unusual translocation of DNA that can interfere with separation of that chromosome pair. Most missing or extra chromosomes cause the resulting embryo to abort even before a pregnancy is detected. An extra copy of chromosome 21 is not as harmful and those embryos survive, resulting in children with Down syndrome. Parents and doctors may carefully consider what choices may be best, but politicians and judges now interfere with their decisions. Fertility clinics have much experience in determining why pregnancies do not occur or do not carry to term.

Screening of all sperm or embryo donors is mandated by the US Food and Drug Administration and the 2024 guidelines require rigorous screening to prevent both genetic and venereal diseases. Thus, assisted conceptions are likely to be less risky and healthier on average than natural conceptions. Sperm donors are the genetic fathers of about 50,000 US children each year. Clinics or national governments may limit how many children may result from one donor, but at least one donor avoided those rules and fathered >550 children across several nations.

Fertility clinics may offer embryo selection for couples with specific genetic defects. Genotypes can be obtained by taking a few cells from very early embryos while still in the lab or from amniotic fluid later in pregnancy. That allows having children without the defect and could allow selecting embryos that inherited chromosomes with more positive effects on health or other traits of value to the potential child. Nature already does much selection at the embryo stage.

Genomics

New tools became available after 2005 to quickly obtain genotypes at thousands of positions across all chromosomes. In 2025, microarrays can read genotypes at 1 million locations for a cost of <$50. Genetic testing using such tools is now a big business in countless species and databases that often include genotypes from millions of individuals per species. Whole genome sequencing can accurately read all 3 billion base pairs of your DNA for a few hundred dollars, with cost decreasing each year. Data growth in the field of genomics often exceeds the rate of gain in computer efficiency. Genomic tools combined with trait data can give people a better understanding of their own DNA and many new options not imagined a few decades ago.

Prediction

Genomic prediction can accurately forecast individual differences for many traits by summing many small genetic effects across all the chromosomes instead of focusing on only a few major genes. Estimating which chromosomes affect traits such as better or poorer health requires very large sets of trait data (phenotypes) matched to DNA data (genotypes). Using many copies of the same chromosomes gives more accurate predictions, and including data from close relatives or from large families helps estimate genetic effects best.

Few businesses or geneticists have helped people find better genes for their children. Fertility clinics offer couples choices on sperm or egg donors. Companies such as 23andMe estimate genetic effects for biologic traits affected by a few major genes such as eye or hair color (see Appendix 1). They provide predictions for some common, useful conditions such as risk of acne, anxiety, asthma, attention deficit disorder, bipolar, depression, some cancers, glaucoma, heart disease, high blood pressure, insomnia, migraines, panic attacks, and sleep apnea, but not for many important traits that you might hope children will have. Newer companies such as LifeView and Nucleus Genomics offer similar services plus embryo testing and selection including a few more traits such as Parkinson’s disease, intelligence quotient, and probability of embryo survival.

What traits are important? Dating sites let people describe their own qualities and those qualities they hope for in a mate. Parents may hope that their children may have some of those same qualities. Lists of qualities to describe yourself or your potential partner are available but many of those qualities are hard to quantify or find data to measure them. Few people describe themselves or their hoped-for date as being free of all the genetic conditions listed above. And they nearly always describe their phenotype (traits they express), not their genotype (their DNA). First impressions and physical appearance may be important in real life. The first internet tool developed by one of the world’s richest men let users rate the faces of young ladies whose pictures he posted. But traits of future children can be predicted more accurately from genotypes than from phenotypes.

Genomic prediction services for humans are not yet as well developed as in agriculture. For millions of cows and bulls, my laboratory at the US Department of Agriculture used genotypes to predict the traits they would transmit to progeny. We closely inspected parent-progeny inheritance and graphed the genetic merit of each animal’s chromosomes. Any cow or bull owner could click on those graphs and see the strengths or weaknesses of each chromosome for each of their traits. Few humans yet know which of their chromosomes have favorable or unfavorable effects on their quality of life or on their specific traits, but they could. Some services can graph which chromosomes you share with your relatives. Some companies may tell you only what percentage of your DNA came from which continent or country.

Advice

Choosing to have kids or not is a big decision that will affect the rest of your life, hopefully in a positive way. Choosing what genes they should have is not as important for you but could be for them. Most children get a random sample of genes from the parents who raised them. Potential parents that are infertile, same-sex, single, or with serious genetic conditions can either adopt or get help to conceive or gestate children. Help for male infertility usually just requires a sperm donor, whereas help for female infertility may require an egg donor, an embryo donor, or a surrogate mother, or embryo adoption with rules more like adopting a child.

Those same methods can help potential parents who for various reasons want to choose the DNA for the children they raise instead of contributing their own DNA. Large fertility clinics select and contract with hundreds of potential donors. Couples often select donors whose face and physical features look like their own to appear more like a natural family, and may use the same donor again to get full rather than half siblings. Some fertility clinics were not monitored closely and one Indiana doctor fathered at least 94 children while lying about using sperm donors, as documented in Our Father.

Donors are tested for many genetic diseases and fertility clinics often provide very detailed phenotypic information about donors. Some donors remain anonymous but more now allow the resulting children to obtain information about their genetic parent, such as after age 18. Because many traits that parents desire may not be highly heritable, genomic predictions could better predict future traits than examining phenotypes of only the donor and the donor’s few close relatives. Some clinics follow up and monitor the resulting children in future years to further assess the genes of the donor. Genomic prediction is not yet widely used in donor selection.

Giving the children you create many years of love and support can be much more important to them than the set of genes they get. Parents can give the same emotional and financial support to assisted children, adopted children, stepchildren, or their own genetic children but that is not guaranteed. Potential parents who cannot conceive often choose other options, whereas those able to have children should balance the potential benefits of using selected DNA vs. their own DNA with the cost, effort, and inconvenience required for assisted methods. Children may prefer their parents’ DNA if the parents’ experience of living with that DNA helps them give better advice and creates a tighter family bond.

Children with DNA selected to help them enjoy life might get more support from parents but they might need less support if they can live their lives without as much help. Only a few generations ago, before Social Security and Medicare, a main benefit of having children was to get support from them when you get old and need help. Children still need much help getting started but then they might spend more time raising the next generation instead of supporting previous generations. Each generation could have better DNA than the previous.

My choices and attitudes

As a graduate student in 1983 I listed quotes on human genetics from many famous leaders and scientists who had noticed that animals and crops were improving each generation and wondered why nobody was improving humans. In 1985 I showed how the same progeny testing methods we used for improving cows could also improve humans, but I did not attempt to publish that paper until finally posting it on my web site in 1999. I did not volunteer to improve humans either. Instead, I took a full-time job improving cows.

I had a girlfriend from Germany for 5 years with interests, attitudes, and education very similar to my own. We coauthored some nice scientific papers on cow genetics. I realized at the time that human genetic improvement would be an even more taboo subject for coauthors with German sounding names, and this topic might distract from my purely agricultural work for the US government. We could not find 2 good jobs together in the same place, we were perhaps too similar, and that relationship ended.

My experience with mate choice is limited but well documented. My wife and I met in 1994 via a dating service where I described my qualities as: Single white male, 33 years old, 6’1”, 175 pounds, solid atheist values, more intelligent than you, sits at home a lot except for tennis, swimming, biking, and basketball, likes Americans, loves foreigners. Hobby: writing easy answers for important questions that affect everyone. Seeking: female, 28-38 years old, any race, any shape except fat, no particular religion, serious, intelligent, and patient. Cheryl was my first arranged date, and she had all the qualities I was seeking, except maybe slightly less patience than I imagined. One year later we were married.

On Valentines Day, 1997 we each gave our one-celled new daughter a random half of our DNA with no attempt to improve her DNA. By pure chance, she got an X instead of Y chromosome from me. For each of her other 22 chromosome pairs, I never checked if she inherited the chromosome copy I got from my Dad, the copy I got from my Mom, or a new, recombinant chromosome that combines some DNA from both of her grandparents in a process called crossing over present in 50% of all chromosomes. In 1997 there were few tools to inspect DNA, and I still have little interest or need to inspect my own DNA at this stage of life. I prefer inspecting cow DNA because that information is very profitable for the global dairy industry and very useful in supplying the nutrients that 8 billion people need.

My wife got a sonogram at 4 months of pregnancy, and we may have started over if the fetus showed major problems of development. Starting over is now illegal in Florida but still legal in Maryland. In 2022 the Supreme Court reversed 50 years of precedent and decided that politicians instead of parents should decide on the future of each embryo.

In recent decades, genetic defects have become much easier to detect but your choices of what to do may be very limited in many states such that pregnancy has become a bigger risk. Many politicians want to make pregnancy even riskier to mothers in all states. For further biological, ethical, and political discussion of that topic see my separate report The Embryos That Do Not Become Babies.

Human genetics may seem a difficult subject, but animal and plant genetic improvement is simple. Just list the traits with data available, predict genetic merit for those traits, estimate how valuable each trait is, multiply each trait’s value by the predicted genetic merit, and sum those together into an index of how desirable that source of DNA is. Then you give the next generation of animals or plants the DNA that will help them live good, productive lives. That was my full-time job for 40 years. The same steps could easily help you choose DNA for your potential child. The traits differ for different species, but the genetic tools and math are almost the same for plants, animals, and humans.

Many people with high moral standards might accuse you of trying to improve the human population. I did not try to improve the DNA of my child, I will have no more children, and I will not be improving or hurting the human population. You can give your child your own DNA like I did, or donate your DNA, or receive other DNA that the next child might enjoy living with. Human genetic options have improved over the decades. Soon, predictions of human traits could become as accurate and useful as trait predictions for crops and livestock already are.

We cannot ask the next generation what DNA they would prefer to receive. Often, parents must make decisions for their young, dependent children. Like farmers, potential parents can think about the options, give the next generation some useful DNA, and let other parents make their choices. You can live a good life while considering how the next child could live an even better life.

References

Gamete and embryo donation guidance (2024) | American Society for Reproductive Medicine | ASRM

CNN SPECIAL REPORT: THE BABY BUSINESS. 2022.

Cole, J.B. and D.J. Null. 2013. Visualization of the transmission of direct genomic values for paternal and maternal chromosomes for 15 traits in US Brown Swiss, Holstein, and Jersey cattle.

Our Father, the fertility doctor who fathered >94 babies. 2022.

The Sperm DONOR that helped 225 families have 550 children

Male infertility crisis - Wikipedia

Everything to know about Facemash, the site Zuckerberg created in college to rank ‘hot’ women

80 Words to Describe Yourself on a Dating Site

Company references

America's Best Fertility Clinics 2025 - Newsweek

Fairfax Cryobank | Find Your Ideal Sperm Donor