Skip to content
BioSyn

Genetics Calculators

Genetics analysis tools

Every tool below opens instantly and runs entirely in your browser — nothing is uploaded, stored or logged.

Inheritance questions look like one tool — draw the square, read the ratio — but the answers people actually need split into different kinds of arithmetic. A square counts zygotes. A family question counts births. A counseling question starts from something already known about a person and has to throw away the outcomes that knowledge excludes. And two genes on the same chromosome stop being a counting problem at all and become a measurement problem. This collection keeps those kinds of arithmetic apart instead of blending them into one ratio generator, and every calculation runs locally in your browser; nothing you type is uploaded.

From Two Known Parents to a Probability

The Punnett square calculator starts where the textbook starts: both parental genotypes are given, and the grid is built by multiplying the gametes each parent can offer. One gene gives four equally likely zygotes, two independently assorting genes give sixteen, and it goes on to four genes. Underneath each percentage the exact box count is printed — nine out of sixteen, not 56.25% floating without a denominator — so you can check a homework answer against the square itself. Complete, incomplete and codominant expression are selectable, which changes only how the boxes are labeled, never how they count. The companion guide to reading a Punnett square walks the margins, the boxes and the point where genotype and phenotype ratios diverge.

The Questions a Grid Cannot Answer

Three inputs on the same page exist because the square is silent on them. Risk across several children treats each birth as an independent draw, so a couple with a 25% per-pregnancy figure and three children faces 57.8% that at least one is affected — two true numbers answering different questions, and mixing them up is the most common error in this whole subject. Conditional probability starts from what is already observed: an unaffected sibling of an affected person is a carrier with probability 2/3, not the 1/2 the raw square suggests, because the affected box no longer belongs in the denominator. The reverse direction — how much two healthy children should lower a couple’s prior carrier chance — is a Bayes update, and the calculator shows the prior, the likelihood and the posterior side by side rather than quietly substituting one for the other.

Sex-linked and linked-gene crosses break the square’s assumptions in two different ways. An X-linked recessive allele lands unequally on sons and daughters, so results are quoted within each sex (50% of sons) and again across pregnancies (25% of children) instead of averaged. Two genes close together on one chromosome stop producing four equal classes; their sizes are set by a recombination fraction you supply, and the deviation is what a map distance in centiMorgans means. The calculator refuses a recombination frequency of 50%, because at that point the cross is indistinguishable from independent assortment and there is no linkage to report.

From a Population Rather Than a Pair

A different question arrives with different evidence: not two parents but a typed sample, or a disease rate, or a screening figure in a leaflet. The Hardy-Weinberg calculator answers that one — allele frequencies counted from genotype classes, the squared proportions a population should show, and the chi-square that decides whether your sample deviates, reported at the degrees of freedom the locus actually has. Beside the test it prints what the usual frequency converters leave out: the inbreeding coefficient F with its sign, so a heterozygote deficit is distinguishable from an excess; the exact test and the sample size a locus needs before either test means anything; a 95% interval on the estimate; and the two inversions people most often confuse, a recessive rate read through q = √(affected) and a dominant one read through the unaffected fraction. X-linked loci are handled as their own mode, because a male rate is q itself while a female rate is q². The companion guide to the equation works through all five directions.

From Inheritance to the Bench

The same alleles show up as physical material in the laboratory, and the quantitation side is already in this suite. DNA concentration converts an absorbance reading into a usable stock, DNA molecular weight works out the mass of a fragment from its length, and GC content summarizes a sequence’s base composition. If the question is which features sit on a molecule and how far apart they are in base pairs — a restriction map, not a genetic map — the restriction site finder answers from the sequence itself. Note that a genetic map distance measured in centiMorgans and a physical one measured in base pairs are different quantities, and the two tools do not convert into each other.

What Is Deliberately Not Here

A cross between two known parents and an allele frequency across a population are different models with different assumptions, and search results routinely collapse them onto one page. This collection keeps them on separate pages: the square predicts a child from genotypes you supply, while the Hardy-Weinberg calculator predicts a sample from frequencies and then tests the fit. Neither is a diagnostic test. Nothing here can establish whether a specific person is a carrier — that comes from assay or pedigree — and none of these numbers should carry a real reproductive decision, which belongs with a genetic counselor who has the family history in front of them. Polygenic traits, mitochondrial inheritance, imprinting and repeat expansions are outside every model on this page.

Frequently asked questions

What can I calculate about inheritance without a test?

Anything that follows from genotypes you already know. If both parents are carriers of the same recessive allele, the Punnett square calculator gives 25% per pregnancy, and from there it can give the chance across a planned family, the carrier probability for a relative whose phenotype you know, and the sex-split risks of an X-linked condition. What it cannot do is tell you whether a specific person is a carrier — that comes from testing or a pedigree, and the calculator takes it as an input.

Is the risk 25% per pregnancy or 25% for the family?

25% per pregnancy, unchanged by how many children you already have. The chance that at least one of three children is affected is 1 − 0.75³ = 57.8%, and the chance that none is affected is 42.2%. Both figures are correct; they answer different questions, and genetic counseling quotes the first while families usually hear the second. The family mode prints both.

Why is an unaffected sibling’s carrier risk 2/3 rather than 1/2?

Because knowing the phenotype removes one of the four boxes. An Aa × Aa cross makes AA, Aa, Aa and aa; if the person is unaffected, aa is no longer a possibility, and two of the three remaining classes are carriers. Conditioning first and then dividing is the general rule — see the step-by-step guide.

Does an X-linked condition come from the mother or the father?

A son receives his single X from his mother and his Y from his father, so an X-linked recessive condition reaches a son through her; a father cannot pass an X-linked allele to a son at all. An affected father’s daughters all inherit his disease X. The X-linked mode reports risks within sons and within daughters separately, plus the figure across all pregnancies, because those denominators are not interchangeable.

When does the 9 : 3 : 3 : 1 ratio stop applying?

As soon as the two genes are close together on the same chromosome. Independent assortment is what makes the four classes equal; linked genes give mostly parental types and few recombinants, with the split set by the recombination fraction. Enter that fraction in the linked-genes mode and the calculator reports parental and recombinant offspring and the map distance implied by it, treating 50% as no detectable linkage.

Is a centiMorgan the same as a physical distance in DNA?

No. One per cent recombination is defined as one centiMorgan, but recombination underestimates physical distance once double crossovers happen between two loci. If you need where features actually sit on a molecule in base pairs, that is a restriction map, and the restriction site finder builds it from the sequence.

How do you turn a disease rate into a carrier frequency?

Take the square root, then double: for a fully penetrant recessive trait the affected are q², so q = √incidence and carriers are 2pq. One birth in 10,000 means q = 1% and carriers at 1.98% — about one person in 50 — which is why a rare recessive allele spends almost all of its time in people who will never have the condition. The Hardy-Weinberg calculator runs that inversion, and it does the dominant version too, where the square root comes from the unaffected fraction instead.

What separates a population calculation from a Punnett square?

The starting point. A square begins with both parents’ genotypes already known and predicts the next child; the squaring rule begins with frequencies across a population and predicts what a sample of strangers looks like, then tests whether your sample matches. Two known parents belong on the Punnett page; a prevalence figure, a typed cohort or a screening percentage belongs on the Hardy-Weinberg page, and keeping the two intents on two pages is why a family risk is never printed on a population tool.

Do these calculators store or upload what I enter?

No. Every calculation runs locally in your browser. The genotypes, probabilities and family sizes you type are never sent to a server, stored or logged, and there is no login.