Veterinary · Pre-clinical and foundation · also MSVE AGB-121
Inheritance from Mendel to molecules, and the population and quantitative genetics that animal breeding is built on.
In the exams: ICAR AICE-JRF core group 05; ICAR AIEEA-PG 14 (AGB); TNPSC AGB 20; ASRB NET (numerical problems, per coaching sources)
Animal genetics is the study of how differences between animals are passed from parents to offspring, and how the frequencies of those differences change in herds, flocks and breeds. It runs from the single gene that makes a calf polled or horned, through the chromosomes that carry genes and the DNA code that genes are written in, to the statistics that let a breeder predict how much of a cow's extra milk her daughters will inherit. This manual covers the principles course (MSVE AGB-121 in the Indian curriculum; the equivalent first genetics course in most veterinary and animal science degrees).
The subject matters in practice for three reasons. First, inherited diseases are real clinical problems: a veterinarian must recognise a recessive defect, advise on carrier testing, and explain risk to an owner in plain numbers. Second, cytogenetic disorders such as the freemartin heifer and the 63,X mare are common causes of infertility that a clinician can confirm with a blood sample. Third, every national breeding programme — progeny testing of bulls, selection of rams, poultry pedigree lines — is applied quantitative genetics, and veterinarians often run or advise these programmes.
Examiners test it in two ways. Recall questions ask for definitions, ratios, chromosome numbers, named defects and the features of the genetic code. Numerical questions, which dominate Indian postgraduate and scientist-recruitment papers (ICAR AIEEA-PG Animal Sciences, ICAR AICE-JRF/SRF core group 05, ASRB NET, state public service commission papers), ask the student to compute Hardy-Weinberg frequencies, changes in gene frequency, inbreeding coefficients, heritability from variance components, and response to selection. Licensing and graduation exams (Nepal NLEV, the Philippine zootechnics paper, Mexico's EGEL reproduction and genetics area) test the same core at a more applied level. Each chapter below ends with the facts and formulas you must be able to use without notes.
Gregor Mendel worked with garden peas and published his results in 1866; they were ignored until 1900, when de Vries, Correns and Tschermak independently reached the same conclusions. Mendel chose true-breeding lines differing in single, clear-cut characters, crossed them, and counted the offspring over several generations. His key insight was that inheritance is particulate: each individual carries two 'factors' (we now say two alleles of a gene) for each character, one from each parent, and these do not blend.
The law of segregation states that the two alleles of a gene separate from each other when gametes form, so each gamete carries only one. A monohybrid cross between two heterozygotes (Aa x Aa) therefore gives genotypes in the ratio 1 AA : 2 Aa : 1 aa and, with complete dominance, phenotypes 3 dominant : 1 recessive. The physical basis is the separation of homologous chromosomes at the first meiotic division.
The law of independent assortment states that alleles of different genes segregate independently of one another. A dihybrid cross (AaBb x AaBb) gives 16 equally likely gamete combinations and a phenotypic ratio of 9 : 3 : 3 : 1. The physical basis is the random orientation of different homologous pairs at metaphase I. The law holds only for genes on different chromosomes or far apart on the same chromosome; genes close together are linked (Chapter 4).
Terms you must use precisely: a gene is a unit of heredity occupying a locus (position) on a chromosome; alleles are alternative forms of a gene; genotype is the allele combination, phenotype the observable trait; homozygous means two identical alleles, heterozygous two different ones; dominant alleles are expressed in the heterozygote, recessive ones only in the homozygote. The F1 is the first filial generation from crossing two parental (P) lines; the F2 comes from intercrossing F1 animals.
Most Mendelian problems are probability problems. Two rules solve nearly all of them. The product rule: the probability that two independent events both occur is the product of their probabilities. The sum rule: the probability that either of two mutually exclusive events occurs is the sum of their probabilities. For example, in cattle polled (P) is dominant to horned (p) and black (B) is dominant to red (b). From BbPp x BbPp, the chance of a red polled calf is 1/4 (bb) x 3/4 (P_) = 3/16.
For a cross heterozygous at n independent loci with complete dominance, the F1 makes 2^n kinds of gamete, the F2 has 3^n genotypes and 2^n phenotypes, and the fraction of F2 homozygous recessive at every locus is (1/4)^n. When you need the chance of a particular mix among several offspring, use the binomial: the probability of exactly k of one kind among n offspring is C(n,k) p^k (1-p)^(n-k).
A test cross mates an individual of dominant phenotype with a homozygous recessive. If any recessive offspring appear, the tested animal must be heterozygous. It is the classical way to test a bull for a recessive defect: mating him to known carrier or affected females and observing the calves. With carrier cows, each calf of a carrier bull has a 1/4 chance of being affected, so the probability of n normal calves by chance is (3/4)^n; 11 normal calves give about 96% confidence (1 - 0.75^11 = 0.958) that the bull is not a carrier, and about 16 give 99%. DNA tests now replace test crosses for most known defects.
A backcross mates the F1 to one of the parental lines. Different results from reciprocal crosses (sexes swapped) point to sex linkage, maternal effects or cytoplasmic inheritance.
Real counts never match expected ratios exactly. The chi-square test asks whether the deviation is larger than chance would explain. For each class compute (observed - expected)^2 / expected and add them; degrees of freedom equal the number of classes minus one. Compare with the table value at 5% (3.84 for 1 df, 5.99 for 2 df, 7.81 for 3 df). If the calculated value is below the table value, the data are consistent with the proposed ratio. For example, 110 black and 50 red calves against an expected 3 : 1 (120 : 40) give chi-square = 100/120 + 100/40 = 3.33, below 3.84, so a 3 : 1 ratio is not rejected.
In animals, controlled matings are expensive, so pedigrees are the everyday evidence. Autosomal recessive traits typically skip generations, appear in offspring of unaffected parents, affect both sexes equally and are more frequent after inbreeding. Autosomal dominant traits appear in every generation and an affected animal usually has an affected parent. X-linked recessive traits affect mostly males, who inherit them from carrier dams.
A frequent clinical calculation: two normal parents produce an affected (recessive) offspring. Both parents are therefore carriers. A normal full sibling of the affected animal has a 2/3 chance of being a carrier, not 1/2, because the homozygous affected class (1/4) is excluded from the normal siblings, leaving 1 AA : 2 Aa.
Dominance describes the interaction of two alleles at one locus. With complete dominance the heterozygote looks like the dominant homozygote. With incomplete (partial) dominance the heterozygote is intermediate; with codominance both alleles are expressed side by side. The roan coat of Shorthorn cattle is the textbook example: red x white gives roan (a mixture of red and white hairs), and roan x roan gives 1 red : 2 roan : 1 white, so phenotype reveals genotype. Blood group antigens and many protein variants detected by electrophoresis are codominant. Overdominance means the heterozygote lies outside the range of both homozygotes.
Lethal alleles kill the carrier, usually during embryonic or early life. Recessive lethals kill only homozygotes. Some alleles are dominant for a visible trait but recessive for lethality, and they distort Mendelian ratios. The yellow mouse allele (Ay), the creeper allele in fowl and the Dexter chondrodysplasia allele in cattle behave this way: heterozygotes show the trait, homozygotes die (in Dexters, the homozygote is the aborted 'bulldog' calf). A cross of two heterozygotes therefore gives 2 affected : 1 normal among live offspring instead of 3 : 1. In horses, foals homozygous for the frame overo allele of the endothelin receptor B gene are born white and die of intestinal aganglionosis (lethal white foal syndrome), while heterozygotes are normally coloured frame overo horses.
Multiple alleles exist when a locus has more than two alleles in the population, though any one diploid animal carries at most two. With n alleles there are n(n+1)/2 genotypes, of which n are homozygous and n(n-1)/2 heterozygous. Coat colour loci (such as the extension and agouti loci) and the major histocompatibility complex are examples. Penetrance is the proportion of animals with a genotype that show the expected phenotype; expressivity is the degree to which it is shown. Pleiotropy is one gene affecting several traits.
Epistasis is the interaction of genes at different loci, where one locus masks or modifies the expression of another. It changes the 9 : 3 : 3 : 1 dihybrid ratio into characteristic modified ratios, and recognising them is a standard exam skill. Recessive epistasis gives 9 : 3 : 4. In Labrador retrievers, the B locus (black B dominant over brown b) is masked when the dog is homozygous ee at the extension locus, which gives yellow regardless of B: BbEe x BbEe gives 9 black : 3 chocolate : 4 yellow.
Dominant epistasis gives 12 : 3 : 1: a dominant allele at one locus masks the other locus. Dominant and recessive (inhibitory) epistasis gives 13 : 3: the classical poultry example is the dominant white inhibitor I, which prevents colour even when the colour allele C is present. A White Leghorn (IICC) crossed with a recessive white breed (iicc) gives white F1 and an F2 of 13 white : 3 coloured, since only iiC_ birds are coloured. Duplicate recessive epistasis (complementary genes) gives 9 : 7, where a dominant allele at both loci is needed for the trait. Duplicate dominant genes give 15 : 1. Interaction without masking gives four phenotypes in 9 : 3 : 3 : 1: comb shape in fowl, where R_P_ gives walnut, R_pp rose, rrP_ pea and rrpp single.
Every modified ratio still adds to 16; only the phenotypic grouping changes.
Most economically important traits — milk yield, growth rate, fleece weight, egg number — are controlled by many loci each with small effect, plus the environment. If loci act additively, each 'plus' allele adds a fixed amount. With n additive loci, an F2 shows 2n + 1 phenotypic classes and the fraction as extreme as one parent is (1/4)^n; if 1/256 of the F2 resemble one parental extreme, four loci are segregating. As loci multiply and the environment blurs the classes, the distribution becomes continuous and approximately normal. Hence quantitative traits are described by means and variances (Chapters 11 and 12).
Non-additive gene action — dominance within loci and epistasis between loci — matters most for crossing. Heterosis (hybrid vigour) is the superiority of crossbred offspring over the average of their parents. Heterosis percentage = (F1 mean - mid-parent mean) / mid-parent mean x 100. It arises mainly from dominance: parental lines fixed for different unfavourable recessive alleles produce F1 offspring that are heterozygous and carry a dominant favourable allele at more loci. Heterosis is largest for low-heritability fitness traits and smallest for carcass traits. Because it depends on heterozygosity, it declines in later generations of inter se mating: under dominance alone, the F2 retains half of the F1 heterosis.
In mammals the female is the homogametic sex (XX) and the male heterogametic (XY). Maleness is triggered by the SRY gene on the Y chromosome, which directs the undifferentiated gonad to become a testis; a Y chromosome therefore makes an animal male even with extra X chromosomes (XXY). In birds the system is reversed: the male is homogametic (ZZ) and the female heterogametic (ZW), so the hen decides the sex of the chick.
Female mammals equalise the dose of X-linked genes by inactivating one X in each somatic cell early in embryonic life (the Lyon hypothesis, proposed by Mary Lyon in 1961). The inactive X is condensed into the Barr body, first seen by Barr and Bertram in the nerve cells of female cats. The number of Barr bodies equals the number of X chromosomes minus one. Inactivation is random and permanent in each cell line, so a heterozygous female is a mosaic of patches expressing one allele or the other.
The tortoiseshell (and calico) cat shows this directly. The orange locus is on the X chromosome: X^O gives orange hair and X^o allows black. A heterozygous female X^O X^o has patches of orange and black; a male, with only one X, is orange or black but not both. A tortoiseshell male is almost always XXY (and usually sterile) or a chimera.
Genes on the X chromosome show criss-cross inheritance: a male receives his X from his dam and passes it to all his daughters but none of his sons. An X-linked recessive trait is expressed in every male that carries it (males are hemizygous) but only in homozygous females, so affected males far outnumber affected females. Haemophilia A (factor VIII deficiency) and haemophilia B (factor IX deficiency) in dogs are X-linked recessive. A carrier bitch mated to a normal dog produces sons of whom half are affected and daughters of whom half are carriers.
In birds, Z-linked genes behave the same way with the sexes reversed. The barring gene (B) of the Barred Plymouth Rock is Z-linked and dominant. Mating a non-barred male (Z^b Z^b) with a barred female (Z^B W) gives barred sons and non-barred daughters, so day-old chicks can be sexed by down colour: this is autosexing, and sex-linked slow and fast feathering is used the same way. Genes on the non-pairing part of the Y chromosome (holandric genes) pass only from father to son.
Sex-linked traits are controlled by genes on the sex chromosomes. Sex-limited traits are controlled by autosomal genes but expressed in one sex only: milk yield, egg production and cryptorchidism. Bulls carry and transmit genes for milk yield although they never lactate, which is why progeny testing is needed. Sex-influenced traits are controlled by autosomal genes whose dominance differs between sexes: in some sheep breeds the horned allele is dominant in rams and recessive in ewes, so heterozygous rams are horned and heterozygous ewes polled. Scurs in cattle are also usually described as sex-influenced.
Genes on the same chromosome tend to be inherited together; they are linked. Linkage was discovered by Bateson and Punnett, and explained by Thomas Hunt Morgan working with Drosophila around 1910-1911. Linked genes are separated only when crossing over occurs between them. Crossing over is the exchange of segments between non-sister chromatids of homologous chromosomes at the pachytene stage of meiotic prophase I; its visible sign is the chiasma, seen at diplotene. Synapsis (pairing of homologues) happens at zygotene.
A dihybrid AaBb can carry its alleles in two arrangements. In coupling (cis) phase the two dominant alleles are on one chromosome (AB/ab); in repulsion (trans) phase each chromosome has one dominant allele (Ab/aB). A test cross reveals the phase: the two most frequent classes are the parental (non-recombinant) types, and the two least frequent are recombinants.
The recombination frequency (RF) is the proportion of recombinant offspring. Because only two of the four chromatids take part in any single crossover, and multiple crossovers can restore the parental arrangement, RF can never exceed 50%; genes far apart on one chromosome assort as if unlinked. For a test cross of an AB/ab animal giving 420 AB, 410 ab, 90 Ab and 80 aB, RF = (90 + 80) / 1000 = 17%, and the genes are 17 map units (centimorgans, cM) apart.
Alfred Sturtevant built the first linkage map in 1913 from the idea that RF is proportional to distance. One map unit (1 cM) equals 1% recombination. In a three-point test cross the rarest classes are double crossovers; comparing them with the parental types identifies the middle gene. Distances are added along the chromosome, and the double-crossover class is counted in both intervals.
One crossover reduces the chance of another nearby; this is interference. Coefficient of coincidence = observed double crossovers / expected double crossovers, where the expected frequency is the product of the two single-interval recombination frequencies. Interference = 1 - coefficient of coincidence. If A-B is 12 cM, B-C is 8 cM and double crossovers are observed at 0.6%, expected doubles = 0.12 x 0.08 = 0.96%, coincidence = 0.625 and interference = 0.375.
Gametes from a known linkage phase are predicted directly from the map distance. An AB/ab animal with the genes 20 cM apart makes 40% AB, 40% ab, 10% Ab and 10% aB gametes. Livestock linkage maps built with microsatellites and SNPs were used to locate quantitative trait loci (QTL). Linkage also explains why a favourable allele can carry an unfavourable neighbour into a population under selection, and why genetic correlations caused by linkage (unlike those caused by pleiotropy) decay over generations.
A chromosome is one long DNA molecule packed with histone proteins into chromatin. At metaphase each chromosome consists of two sister chromatids joined at the centromere. By centromere position chromosomes are metacentric (central), submetacentric, acrocentric (near one end) or telocentric (at the end). The karyotype is the chromosome complement of a species, arranged by size and shape.
Diploid numbers (2n) that every student must know: cattle 60, goat 60, sheep 54, river buffalo 50, swamp buffalo 48, horse 64, donkey 62, pig 38, dog 78, cat 38, rabbit 44, chicken 78, human 46. In cattle all 29 pairs of autosomes are acrocentric, and the X and Y are the only bi-armed chromosomes; this makes cattle karyotypes simple to screen for fusions. The chicken karyotype has a few large macrochromosomes and many tiny microchromosomes. The mule (mare x donkey jack) has 63 chromosomes and is almost always sterile because its unpaired chromosomes disrupt meiosis.
Chromosomes are studied in dividing cells. The routine method cultures peripheral blood lymphocytes stimulated with a mitogen, arrests them at metaphase with colchicine, swells them in hypotonic solution, fixes them in methanol-acetic acid and stains them. Banding (for example G-banding with trypsin and Giemsa) gives each chromosome a specific pattern so that homologues and rearrangements can be identified.
Numerical abnormalities are changes in chromosome number. Euploidy changes involve whole sets (triploidy, 3n; tetraploidy, 4n); polyploidy is common in plants but lethal in mammals and is seen mainly in early embryonic loss. Aneuploidy is the gain or loss of single chromosomes: monosomy (2n - 1), trisomy (2n + 1). It arises from non-disjunction — failure of homologues or sister chromatids to separate at meiosis or mitosis. Sex-chromosome aneuploidies (XO, XXY, XXX) are better tolerated than autosomal ones.
Structural abnormalities include deletions, duplications, inversions (a segment reversed) and translocations (a segment moved to a non-homologous chromosome). A reciprocal translocation exchanges segments between two chromosomes. A Robertsonian translocation (centric fusion) joins two acrocentric chromosomes at their centromeres into one bi-armed chromosome, reducing the count by one. Carriers are usually phenotypically normal because no genetic material is lost, but they produce unbalanced gametes, which causes early embryonic death and reduced fertility.
The 1/29 Robertsonian translocation of cattle, first described by Gustavsson in the Swedish Red and White breed, is the best known chromosome abnormality in livestock. It occurs in many breeds worldwide and causes a modest but economically important fall in fertility of carrier animals, which is why many AI centres karyotype young bulls before use. Mosaicism means two or more cell lines of different chromosome make-up arising in one zygote; chimerism means cell lines derived from different zygotes.
The freemartin is the commonest clinical example. When a heifer is born co-twin to a bull, the placental circulations usually fuse early in pregnancy, allowing exchange of blood-forming cells and hormones. The heifer becomes an XX/XY blood chimera and her reproductive tract is masculinised to a varying degree: a short blind vagina, absent or rudimentary cervix and uterus, and gonads that may resemble testes. The great majority of such heifers — commonly quoted as over 90% — are freemartins and are sterile; the bull co-twin is usually fertile or has only slightly reduced fertility. Diagnosis is by measuring vaginal length compared with a normal heifer of the same age, and confirmed by karyotyping blood lymphocytes for XX/XY chimerism or by PCR for Y-chromosome-specific sequences. Freemartinism also occurs in sheep, goats and pigs, though less often.
In the horse, 63,X gonadal dysgenesis (the equivalent of human Turner syndrome) is the most frequently reported chromosomal cause of infertility in mares: affected mares are often small, with small inactive ovaries and a flaccid uterus, and never cycle normally. XY sex reversal (an XY karyotype with a female phenotype) is also reported in mares. In cats, XXY males are tortoiseshell and usually sterile. The practical rule: any female with a persistently abnormal tract or no cycles, and any heifer twin to a bull, should be karyotyped or tested for Y sequences before money is spent on breeding her.
A mutation is a heritable change in the DNA sequence. Mutations in germ-line cells can be passed to offspring; somatic mutations affect only the individual and are the basis of most cancers. Mutations are the ultimate source of genetic variation.
Point mutations change a single base. A transition replaces a purine with a purine (A to G) or a pyrimidine with a pyrimidine (C to T); a transversion swaps a purine for a pyrimidine or the reverse. By their effect on the protein, point mutations are silent (same amino acid, possible because the code is degenerate), missense (different amino acid), or nonsense (a stop codon, giving a truncated protein). Insertions or deletions of bases that are not multiples of three cause frameshift mutations, which change every codon downstream and usually produce a non-functional protein; insertions or deletions of three bases remove or add an amino acid but keep the reading frame. Larger changes — deletions, duplications and rearrangements of chromosome segments — were covered in Chapter 5.
Mutations arise spontaneously from replication errors, tautomeric shifts of bases and spontaneous chemical changes such as deamination of cytosine to uracil. Mutagens raise the rate. Base analogues such as 5-bromouracil are incorporated in place of normal bases and cause transitions; alkylating agents modify bases; acridine dyes intercalate between bases and cause frameshifts; ultraviolet light produces thymine dimers; ionising radiation causes strand breaks. Classical estimates of spontaneous mutation rates are roughly one per 100,000 to one per million gametes per locus per generation, varying widely by locus.
Many single-gene defects of domestic animals have been traced to specific mutations, and DNA tests for them are routine in breeding programmes. Bovine leukocyte adhesion deficiency (BLAD) in Holstein cattle is an autosomal recessive caused by a single-base missense mutation in the ITGB2 gene, which codes for the CD18 integrin; affected calves have recurrent infections because neutrophils cannot leave the blood vessels. Complex vertebral malformation (CVM) in Holsteins is an autosomal recessive caused by a mutation in the SLC35A3 gene, resulting in abortion and malformed calves. Deficiency of uridine monophosphate synthase (DUMPS) and bovine citrullinaemia are other recessive defects with known mutations.
Porcine stress syndrome (malignant hyperthermia, the halothane gene) is caused by a point mutation in the ryanodine receptor gene RYR1, which controls calcium release in muscle. Homozygous pigs are leaner and more heavily muscled but can die of stress and produce pale, soft, exudative (PSE) meat. Hyperkalaemic periodic paralysis (HYPP) in Quarter Horses is an autosomal dominant mutation in the sodium channel gene SCN4A. Double muscling in cattle breeds such as the Belgian Blue results from loss-of-function mutations in the myostatin (MSTN) gene, which normally limits muscle growth; the Belgian Blue carries an 11-base-pair deletion that causes a frameshift. Susceptibility to classical scrapie in sheep depends on variants at codons 136, 154 and 171 of the prion protein gene PRNP; the ARR allele confers the most resistance and VRQ the most susceptibility, which formed the basis of scrapie-resistance breeding schemes.
Online Mendelian Inheritance in Animals (OMIA) catalogues these traits and their causal variants by species.
Watson and Crick proposed the double helix in 1953. DNA is a polymer of nucleotides, each made of a deoxyribose sugar, a phosphate and a base: the purines adenine (A) and guanine (G), and the pyrimidines cytosine (C) and thymine (T). Two antiparallel strands are held by hydrogen bonds between complementary bases: A pairs with T by two hydrogen bonds, G with C by three. Chargaff's rules follow: A = T, G = C, and purines = pyrimidines. So if a double-stranded DNA is 30% A, it is also 30% T, leaving 40% for G + C, or 20% each. DNA with more G-C pairs has a higher melting temperature. In the common B form, the helix is right-handed, about 2 nm wide, with about 10 base pairs per turn and 0.34 nm between base pairs (3.4 nm per turn). RNA differs by having ribose and uracil instead of thymine, and is usually single stranded.
Replication is semi-conservative: each daughter molecule keeps one parental strand, as Meselson and Stahl showed in 1958 by growing bacteria in heavy nitrogen (15N) and then light nitrogen (14N). After one generation in 14N all DNA is hybrid; after two, half is hybrid and half light; after n generations the hybrid fraction is 2/2^n. DNA polymerases add nucleotides only to a 3'-OH end, so synthesis runs 5' to 3'. The leading strand is made continuously; the lagging strand is made in short Okazaki fragments, each started by an RNA primer from primase and later joined by DNA ligase. Helicase unwinds the helix.
The central dogma describes information flow from DNA to RNA to protein (with reverse transcription from RNA to DNA in retroviruses). In transcription, RNA polymerase reads the template (antisense) strand 3' to 5' and makes mRNA 5' to 3', so the mRNA has the same sequence as the coding (sense) strand, with U for T. Eukaryotic genes contain exons, which are kept, and introns, which are removed by splicing; the primary transcript also receives a 5' cap and a 3' poly-A tail. In the bacterial lac operon, a repressor binds the operator and blocks transcription until the inducer allolactose releases it.
Translation occurs on ribosomes. Transfer RNAs carry amino acids and recognise codons with their anticodons. The genetic code, deciphered in the early 1960s by Nirenberg, Matthaei, Khorana and others (Nirenberg and Matthaei's poly-U RNA gave polyphenylalanine), has these features: it is a triplet code; 4^3 = 64 codons exist, 61 specify the 20 amino acids and 3 are stop codons (UAA, UAG, UGA); AUG codes methionine and is the usual start codon; the code is degenerate (most amino acids have more than one codon), non-overlapping and without punctuation (commaless); and it is nearly universal. Exceptions exist mainly in mitochondria: in vertebrate mitochondria, UGA codes tryptophan instead of stop. Crick's wobble hypothesis explains why one tRNA can read several codons: pairing at the third codon position (the anticodon's first base) is less strict.
A coding sequence of n nucleotides including the stop codon specifies (n/3) - 1 amino acids. Epigenetic marks such as DNA methylation and histone modification alter expression without changing the sequence.
Karyotyping (Chapter 5) remains the tool for detecting chromosome number changes, translocations and chimerism. Fluorescence in situ hybridisation (FISH) uses a fluorescently labelled DNA probe that binds its complementary sequence on chromosomes or in nuclei, so that a specific gene, chromosome or rearrangement can be seen under the microscope; it is used to confirm translocations and to identify sex chromosomes.
DNA is extracted from blood, hair roots, semen, tissue or ear-notch samples. Restriction endonucleases cut DNA at specific short recognition sequences, often palindromes. A random sequence contains a given 4-base recognition site on average once in every 4^4 = 256 base pairs, and a 6-base site once in 4^6 = 4,096 base pairs. In agarose gel electrophoresis DNA, which is negatively charged by its phosphate backbone, moves toward the positive electrode (anode); smaller fragments move faster and further, and sizes are read against a ladder of known fragments. Southern blotting transfers DNA from a gel to a membrane for probing (Northern for RNA, Western for protein).
The polymerase chain reaction (PCR), developed by Kary Mullis in the mid-1980s, copies a chosen DNA segment exponentially. Each cycle has three steps: denaturation (about 94-95 C) separates the strands; annealing (typically 50-65 C, set by the primers) lets two short primers bind either side of the target; and extension (72 C) lets a heat-stable DNA polymerase from Thermus aquaticus (Taq) copy the target. In theory the number of copies doubles each cycle, giving 2^n copies after n cycles from one template (about one billion after 30 cycles); in practice efficiency falls in late cycles. A rough rule for the melting temperature of a short primer (the Wallace rule) is Tm = 2(A + T) + 4(G + C) degrees Celsius. Real-time PCR measures product as it accumulates; reverse-transcription PCR starts from RNA.
PCR-RFLP (restriction fragment length polymorphism) is the classic way to genotype a known point mutation: amplify the region, cut it with an enzyme whose site is created or destroyed by the mutation, and read the fragment pattern on a gel. Tests for BLAD, the halothane gene and many coat-colour alleles were first run this way. Allele-specific PCR and TaqMan-type probe assays do the same job more quickly.
Sanger sequencing (1977) uses dideoxynucleotides, which lack the 3'-OH group and so stop chain extension, producing fragments that end at each base; capillary electrophoresis reads the sequence.
Microsatellites (short tandem repeats) are highly polymorphic codominant markers and were the international standard for parentage verification and identity in cattle, horses, dogs and other species, using panels recommended by the International Society for Animal Genetics (ISAG). Parentage exclusion rests on Mendel's first law: a true parent must share at least one allele with the offspring at every locus (barring mutation or typing error), and the calf's other allele must come from the other parent. SNP (single nucleotide polymorphism) chips, which genotype tens of thousands of markers at once, have largely replaced microsatellites for parentage in cattle and are the basis of genomic selection.
Population genetics describes a group of interbreeding animals by the frequencies of alleles and genotypes rather than by individual crosses. For a locus with alleles A and a, write P, H and Q for the frequencies of AA, Aa and aa. The allele frequencies are p = P + H/2 and q = Q + H/2, with p + q = 1. When genotypes can be counted — codominant traits, blood groups, DNA markers — the allele frequency is simply the number of copies of the allele divided by twice the number of animals. In a sample of 360 AA, 480 Aa and 160 aa (1,000 animals, 2,000 alleles), p = (720 + 480) / 2000 = 0.6.
The Hardy-Weinberg law, stated independently by G. H. Hardy and Wilhelm Weinberg in 1908, says that in a large, randomly mating population with no selection, mutation, migration or drift, allele frequencies stay constant from generation to generation, and genotype frequencies reach p^2, 2pq and q^2 after a single generation of random mating (for an autosomal locus with equal allele frequencies in both sexes). Random mating of genotypes is equivalent to random union of gametes, which is why the binomial expansion (p + q)^2 gives the answer. The same logic extends to multiple alleles: with frequencies p, q and r, genotype frequencies are the terms of (p + q + r)^2, and the total heterozygosity is 1 minus the sum of the squared allele frequencies.
Two useful consequences: heterozygosity 2pq is greatest (0.5) when p = q = 0.5; and when a recessive allele is rare, almost all copies of it are hidden in carriers. The ratio of carriers to affected animals is 2pq / q^2 = 2p/q, which is about 198 when q = 0.01.
The most common applied calculation estimates carrier frequency from the incidence of a recessive defect. If 1 calf in 400 is affected, q^2 = 0.0025, q = 0.05, p = 0.95 and the carrier frequency is 2pq = 0.095, or about 1 animal in 10.5. This assumes random mating with respect to the locus; inbreeding or popular-sire effects make the true carrier frequency locally higher.
For a dominant trait, the recessive phenotype frequency is still q^2. If 91% of a herd is polled (dominant), 9% are horned, so q = 0.3 and p = 0.7. The proportion of polled animals that are homozygous PP is p^2 / (p^2 + 2pq) = 0.49 / 0.91, about 54%. For a codominant trait such as roan in Shorthorns, count directly: a herd of 49% red, 42% roan and 9% white has a white allele frequency of 0.09 + 0.42/2 = 0.30.
To test whether a sample is in equilibrium, compute allele frequencies from the sample, compute expected genotype numbers, and apply chi-square. For two alleles there are three genotype classes but one degree of freedom, because one df is lost for the total and one for estimating p from the data. For 500 AA, 200 Aa and 300 aa, p = 0.6, expected numbers are 360, 480 and 160, and chi-square = 140^2/360 + 280^2/480 + 140^2/160 = 340, far above 3.84: a large heterozygote deficit, suggesting inbreeding, population subdivision (the Wahlund effect) or selection against heterozygotes.
For an X-linked locus, male genotype frequencies equal allele frequencies (p and q), and female genotype frequencies are p^2, 2pq and q^2. If 20% of male cats are orange, then among females 2pq = 2 x 0.8 x 0.2 = 0.32 are tortoiseshell and 0.04 are orange.
Four forces change allele frequencies. Migration, mutation and selection are systematic: they change frequency by a predictable amount and direction. Random genetic drift is dispersive: its direction is unpredictable, but its size can be predicted.
Migration (gene flow). If a proportion m of breeding animals each generation are immigrants with allele frequency qm, the new frequency is q1 = (1 - m)q0 + m qm, so the change is delta q = m(qm - q0). Grading up a local breed with imported semen is migration at a high rate. Mutation. If A mutates to a at rate u and a back to A at rate v, delta q = up - vq, and at equilibrium q = u / (u + v). Because rates are small, mutation alone changes frequencies extremely slowly.
Selection. Fitness is the relative contribution of a genotype to the next generation; the selection coefficient s is the proportional reduction in fitness compared with the best genotype. With complete dominance and selection against the recessive homozygote (fitnesses 1, 1, 1 - s), delta q = -s q^2 (1 - q) / (1 - s q^2). When aa is lethal or never allowed to breed (s = 1), q1 = q0 / (1 + q0) and the number of generations to fall from q0 to qt is t = 1/qt - 1/q0. From q = 0.1 to q = 0.01 takes 90 generations — which is why culling affected animals alone removes a recessive defect very slowly, and why carrier testing is needed. Selection against a dominant allele, by contrast, removes it in one generation if the dominant phenotype is fully penetrant and completely culled. When the heterozygote is fittest (overdominance, fitnesses 1 - s, 1, 1 - t), a stable equilibrium results with q = s / (s + t).
In a finite population each generation is a sample of the gametes of the last, and allele frequencies change by chance. The variance of the change in one generation is pq / 2N. Over time drift leads to fixation or loss of alleles, differentiation between isolated lines, and loss of heterozygosity. Drift is strongest in small populations: rare breeds, closed nucleus herds, zoo populations and lines kept for research.
The effective population size (Ne) is the size of an idealised population that would drift at the same rate as the real one. With unequal numbers of breeding males (Nm) and females (Nf), Ne = 4 Nm Nf / (Nm + Nf); a herd of 10 bulls and 200 cows has Ne of only about 38, dominated by the few males. Across generations Ne is the harmonic mean of the sizes, so bottlenecks dominate. The rate of inbreeding per generation is delta F = 1 / (2Ne), and after t generations F = 1 - (1 - delta F)^t. A widely used rule of thumb in conservation is to keep delta F at or below about 1% per generation, which corresponds to Ne of at least 50.
Inbreeding is the mating of animals more closely related than the average of the population. Sewall Wright's inbreeding coefficient F is the probability that the two alleles at a locus in an individual are identical by descent (copies of the same ancestral allele). It is computed from a pedigree by path counting: F_X = sum over all paths through common ancestors of (1/2)^n (1 + F_A), where n is the number of individuals in the path joining the sire and dam through the common ancestor A (counting both parents and the ancestor), and F_A is the ancestor's own inbreeding coefficient. Offspring of a half-sib mating have F = (1/2)^3 = 0.125 if the common ancestor is not inbred, and 0.1875 if that ancestor has F = 0.5. Full-sib and parent-offspring matings give F = 0.25; first cousins give 0.0625. The coefficient of relationship (Wright's additive relationship between two non-inbred animals) is 0.5 for full sibs and parent-offspring, 0.25 for half sibs and 0.125 for first cousins; an animal's inbreeding coefficient is half the relationship between its parents.
Inbreeding does not change allele frequencies by itself, but it redistributes genotypes: homozygote frequency rises by Fpq for each homozygote class and heterozygosity falls to H0(1 - F). For a recessive allele at q = 0.1, the affected frequency rises from q^2 = 0.01 under random mating to q^2 + Fpq = 0.0325 at F = 0.25, more than three times higher. This exposure of recessive defects, together with loss of favourable dominance effects, causes inbreeding depression — a decline in fitness traits such as fertility, survival, litter size and growth, roughly proportional to F. Inbreeding is still used deliberately to fix type and to build lines for crossing; linebreeding concentrates the genes of an outstanding ancestor while keeping F moderate.
A quantitative trait is measured, not classified. The phenotypic value of an animal is P = G + E, where G is the genotypic value and E the environmental deviation. The genotypic value splits into the breeding value A (the sum of the average effects of the animal's alleles), the dominance deviation D (interaction of alleles at a locus) and the epistatic (interaction) deviation I. Correspondingly the phenotypic variance is VP = VA + VD + VI + VE, assuming genotype and environment are uncorrelated and do not interact.
For a single locus with homozygote values +a and -a and heterozygote value d, the population mean is M = a(p - q) + 2dpq. The average effect of substituting A for a is alpha = a + d(q - p). The additive variance is VA = 2pq alpha^2 and the dominance variance VD = (2pqd)^2. These formulas (Falconer and Mackay) show that additive variance depends on allele frequency and on dominance: an allele that is completely dominant contributes little additive variance once it is common.
Breeding value is the value of an animal as a parent: twice the expected deviation of its offspring from the population mean when it is mated at random. Only half of an animal's breeding value is passed on to each offspring; that half is called the transmitting ability (and in dairy cattle, predicted transmitting ability, PTA). Dominance and epistatic deviations are not transmitted because parents pass on alleles, not genotypes.
Heritability in the broad sense is H^2 = VG / VP. Heritability in the narrow sense, which is what breeders mean by 'heritability', is h^2 = VA / VP: the proportion of phenotypic variance due to breeding values, and equally the regression of breeding value on phenotype. It ranges from 0 to 1 and belongs to a particular trait in a particular population and environment; it is not a fixed property of the trait. As a rough guide, fitness traits such as fertility, litter size and survival have low heritability (below about 0.15); milk yield and growth rate are moderate (about 0.2 to 0.4); carcass and body composition traits, milk fat percentage and mature size are high (above about 0.4).
Heritability is estimated from the resemblance between relatives. For offspring regressed on one parent, b = h^2/2, so h^2 = 2b; on the mid-parent (average of both parents), h^2 = b. For paternal half sibs, the intraclass correlation t = h^2/4, so h^2 = 4t; in an analysis of variance of sire families, h^2 = 4 sigma^2_s / (sigma^2_s + sigma^2_w), where sigma^2_s is the between-sire and sigma^2_w the within-sire variance component. Full-sib correlations contain dominance and common-environment effects (for example the shared uterus and litter), so they overestimate h^2. Realised heritability is the ratio of cumulative response to cumulative selection differential in a selection experiment.
Heritability tells a breeder how reliably an animal's own performance indicates its breeding value. The estimated breeding value from one own record is EBV = h^2 (P - population mean). A cow producing 1,000 kg above her herd-mates with h^2 = 0.25 has an EBV of +250 kg and a transmitting ability of +125 kg. For a sire judged on the mean of n half-sib progeny, the regression is b = 2nh^2 / (4 + (n - 1)h^2) and the accuracy is the square root of nh^2 / (4 + (n - 1)h^2). With 40 daughters and h^2 = 0.25, accuracy is about 0.85 — which is why progeny testing gives far more accurate proofs than a bull's own records could for a sex-limited trait.
Many traits are measured more than once in an animal's life: milk yield in successive lactations, litter size in successive farrowings, fleece weight each year. Repeatability (r) is the correlation between repeated records of the same animal. It divides the environmental variance into a permanent part (VEp: effects that persist through life, such as early nutrition or loss of a quarter of the udder) and a temporary part (VEt: effects specific to one record). Then r = (VG + VEp) / VP. Repeatability is estimated as the intraclass correlation from an analysis of variance of repeated records: r = sigma^2_between animals / (sigma^2_between + sigma^2_within).
Because VG includes VA, repeatability sets an upper limit to both broad and narrow heritability: h^2 cannot exceed r. Lactation milk yield is usually moderately to highly repeatable, while fertility traits such as calving interval are much less repeatable; exact values depend on the population. Repeatability has three uses. It predicts future performance: the most probable producing ability of a cow with n records averaging a deviation P from the herd mean is MPPA = [nr / (1 + (n - 1)r)] x P. It measures the gain from taking more records: the heritability of the mean of n records is nh^2 / (1 + (n - 1)r), which rises with n most when r is low. And it guides culling: when r is high, the first record is a good guide to later ones and early culling is efficient.
Two traits are genetically correlated when the same genes affect both. The genetic correlation rA is the correlation between the breeding values of the two traits: rA = cov_A(x,y) / (sigma_Ax sigma_Ay). Its causes are pleiotropy, which is permanent, and linkage disequilibrium, which is transient and decays with recombination. The phenotypic correlation combines genetic and environmental correlations weighted by the square roots of heritability and environmentality: rP = hx hy rA + ex ey rE, where e^2 = 1 - h^2. The phenotypic and genetic correlations can therefore differ in size and even sign.
Genetic correlations matter because selection for one trait changes others. The correlated response in trait y to selection on trait x is CRy = i hx hy rA sigma_Py. Selecting on a correlated trait can beat direct selection when the correlated trait has higher heritability or can be measured earlier, more cheaply or in both sexes: the relative efficiency of indirect to direct selection (same intensity and interval) is rA hy / hx. Important unfavourable genetic correlations in livestock include that between milk yield and fertility in dairy cattle and between milk yield and fat or protein percentage, which is why modern breeding goals combine several traits in a selection index.
The selection differential S is the mean superiority of the selected parents over the population mean. The response R — the change in the mean of the offspring generation — is R = h^2 S. Because S depends on the unit of measurement, it is standardised: S = i sigma_P, where i is the intensity of selection, the mean deviation of the selected group in standard deviation units. For truncation selection from a normal distribution, i is about 0.80 when the best 50% are kept, 1.40 for 20%, 1.76 for 10%, 2.06 for 5% and 2.67 for 1%. Hence R = i h^2 sigma_P = i h sigma_A, which shows the four levers a breeder controls: intensity, accuracy (h, or the accuracy of an EBV), genetic variation (sigma_A) and, per year, generation interval.
The generation interval (L) is the average age of parents when their offspring are born. Annual response is R per generation divided by L. When males and females are selected differently, the Rendel-Robertson form gives annual genetic gain = (im + if) h^2 sigma_P / (Lm + Lf) for selection on own phenotype, or more generally the sum of the gains through each path divided by the sum of the intervals. Males are usually selected far more intensely than females because one sire, especially with artificial insemination, mates many dams. For example, with top 5% of males (i = 2.06), top 50% of females (i = 0.80), h^2 = 0.3, sigma_P = 20 kg and intervals of 3 and 5 years, annual gain is about 2.1 kg.
Response predicted by R = h^2 S holds for one or a few generations. Over longer periods, response may slow as favourable alleles approach fixation, as inbreeding reduces variance, or as natural selection opposes artificial selection; a selection plateau may be reached.
A first-year student uses this subject to solve genetics problems quickly and correctly: set up a Punnett square or product-rule calculation, recognise a modified dihybrid ratio, compute Hardy-Weinberg carrier frequencies, and trace an inbreeding coefficient through a pedigree. The same skills carry the numerical questions in postgraduate entrance and scientist-recruitment papers. Learn the formulas in the key points until you can apply them without notes, and always write the working: most lost marks come from using q instead of q^2, forgetting to double a regression, or mixing selection differential with response.
A practitioner five years out meets genetics as clinical questions. A client's heifer born co-twin to a bull must be tested for freemartinism before she is bred; a mare that never cycles should be karyotyped. A litter with a recessive defect means both parents are carriers, and the practitioner must explain that normal littermates have a 2/3 chance of carrying it and advise DNA testing before they are bred. Dairy and pig clients ask about carrier status for BLAD, CVM or the halothane gene on bulls and boars, and the vet must read a test report, advise on mating carriers to non-carriers, and explain why culling affected animals alone barely lowers the allele frequency.
A veterinarian with 25 years in field service or a breeding organisation designs and audits breeding programmes. The decisions are quantitative: how many bulls to progeny test and with how many daughters to reach a target accuracy, how to balance selection intensity against inbreeding when semen from a few elite sires dominates, whether a local breed's effective population size is falling below safe limits, and how to handle an emerging recessive defect traced to a popular sire. They must also explain to farmers and policy makers what a heritability or breeding value means, and what it does not.
A researcher with a PhD and long experience works at the edges of this theory: estimating variance components and genetic correlations with mixed models and animal-model REML, mapping causal variants with genome-wide association and sequence data, modelling genotype-by-environment interaction as breeds move between climates, and managing genetic diversity in conservation programmes using genomic relationships. The classical results in this manual remain the reference points: the infinitesimal model, the Hardy-Weinberg baseline, Wright's F and Ne, and the breeder's equation. Much published work tests where these approximations fail, so a researcher must know exactly which assumptions each formula carries.
This manual covers principles only. Selection methods and indices, mating systems and crossbreeding designs, sire evaluation models (BLUP, animal model), genomic selection, national breeding policies and conservation of animal genetic resources belong to the livestock and poultry breeding course. Molecular genetics is covered only as far as the genetic code, gene expression basics and routine laboratory techniques; bacterial genetics, recombinant DNA technology and transgenesis are left to the biotechnology course. Heritability and repeatability values are given only as broad classes, because published estimates vary by breed, population, environment and method. Named mutations are limited to well-established examples; the list of inherited disorders for any breed should be checked in OMIA and breed-society records, which change as new variants are found.
Original practice questions written to the syllabus and pattern of the exams named in each question's examTags. They are not past papers and do not reproduce any board's questions. Every numerical answer was computed by script, and the working is shown in the explanation.
These notes are free to read here, with no account.
What the Laisha app adds:
Free for ten days, then a subscription. We would rather write that here than have you find it later.