I don't have a background in biology, but is trying to learn more about genetics; I have watched many videos about SNPs and still feel confused about the concept of a single nucleotide polymorphism. According to wikipedia, a SNP "is a germline substitution of a single nucleotide at a specific position in the genome". What I am most puzzled by is the concept of a 'specific position in the genome', if chromosome lengths differ, how can we talk of a 'specific position'?

For example, if we have two arrays of items that are of the same length:[A,C,C,C,A,T,G] and [A,C,G,C,A,T,G] then it is straightforward to that the entries at the third position of each array are different. But with chromosomes, their lengths differ, so it might look like: [A,C,C,C,A,T,G] and [A,C,C,C,A,A,T,G], now how can we say what corresponds to what? Do we say there is a SNP at the fifth/sixth base pair?

What about something more different like [A,C,C,C,A,A,T,G] and [C,C,A,A]? How do we know what corresponds to what? What are the SNPs?

What if two individuals each has a chromosome 1 that differ by many base pairs, how do we know what corresponds to what?


3 Answers 3


The reason we can compare SNPs between different people is because sequencing reads are aligned to a reference genome with a fixed set of coordinates for each chromosome (local alignment), rather than aligning the entire chromosomes of two different people (global alignment). For instance, 10:3424234 refers to the 3424234th base-pair on chromosome 10 on the reference genome.

Say we wish to compare this particular SNP between two different people; we can sequence both individuals and reads covering the 3424234th base-pair on chromosome 10 will be produced and aligned to the same part of the reference genome; we are certain that the reads originate from the same part of the genome in both individuals. Then, it's trivial to look at this position and see whether it differs between the two individuals. It therefore doesn't matter if one individual has chromosome 10 twice the length of the other (although this would never be the case in reality), as long as the read originates from the same part of the genome in both individuals, they can be compared.

enter image description here

  • 1
    $\begingroup$ In the image, say that we are interested in base pair 114. Read2 is aligned to that position and tells us the genotype at this position. If we do this for two people, then we can compare the sequences on the reads to see if they are the same or different. Does that help? $\endgroup$
    – user438383
    Commented Jan 20, 2022 at 17:38
  • 1
    $\begingroup$ Please allow me to ask 2 different questions: How do we know that Read1 is correctly aligned? There are 2 discrepancies. How do we know that Read3 is correctly aligned? Is this all done by statistical/computational methods? $\endgroup$ Commented Jan 20, 2022 at 17:43
  • 1
    $\begingroup$ Yes, the reads are about 25 base-pairs long, and there are sophisticated algorithms to make sure they are aligned correctly, e.g. ncbi.nlm.nih.gov/pmc/articles/PMC2705234 $\endgroup$
    – user438383
    Commented Jan 20, 2022 at 17:44
  • 1
    $\begingroup$ Read3 has two extra C's in the middle, does this imply this person inherited a mutation (or two mutations), specifically, 2 C insertions? It looks like this procedure works because human genomes are largely the same. $\endgroup$ Commented Jan 20, 2022 at 17:49
  • 2
    $\begingroup$ Yes - exactly right on both counts. $\endgroup$
    – user438383
    Commented Jan 20, 2022 at 17:54

It may be easier to understand this with an analogy. Think of the genome as a book, and the chromosome as a chapter in the book.

If someone were copying the book by hand (as was common before the invention of the printing press), they'll make some mistakes. The vast majority will be copied correctly, but here and there they'll misspell something, and might even occasionally leave out of duplicate a word.

These mistakes are analogous to mutations during genome replication. As long as they're rare, you can easily recognize all the common parts of the original and copy, and the places with the errors stand out as SNPs.

E.g. if you go from

The quick brown fox jumped over the lazy dog


The quicck brown fox jumped over a lazy dog

you can see that an extra "c" was added in the second word, and the second "the" was replaced with "a". But for the most part the two sentences are the same. And just as we're able to make sense of the sentence with these errors, the genome has enough redundancy (as well as "junk" that has little effect) that the genome still works with a few errors and might even be beneficial (but some can cause diseases, or be catastrophic enough to prevent the organism from developing at all).

SNPs aren't the only kinds of mutations, they're just some of the easiest to analyze because they stand out so well. For instance, humans have 46 chromosomes, while great apes have 48. Scientists have determined that human chromosome 2 resulted from the fusion of two chromosomes from the common ancestor of humans and apes. The analogy would be an editor deciding to combine chapters of the book when releasing a new edition (although in evolution there's no editor making conscious decisions, it just happened accidentally, as if the chapter heading was left out when copying). It took some genetic sleuthing to determine whether this was originally two chromosomes that merged in humans, or one chromosome that split in apes; you can read about it here.

  • $\begingroup$ Agreed this is a good analogy. Some other aspects that the analogy helps with is showing why you need context; if you just had the word "quick" it would probably be hard to ID the typo as just a "SNP", just like in OP's 4-7 base examples (maybe the original was "duck" and both the q and i are inserted). But, once you have the whole sentence, it's clear where exactly the difference lies. It also doesn't matter if you only started copying from Chapter 3, or have extra introductory pages; it's still clear where the exact change is from the local context, regardless of the total length. $\endgroup$
    – Bryan Krause
    Commented Jan 21, 2022 at 16:19
  • $\begingroup$ Good analogy, thank you. $\endgroup$ Commented Jan 21, 2022 at 19:17
  • $\begingroup$ But as more and more mutations occur (like in 500 million years), the genomes will become less and less comparable, right? $\endgroup$ Commented Jan 22, 2022 at 9:26
  • $\begingroup$ @NoppaweeApichonpongpan Yes, although there are still meaningful similarities. There are some genes that control some basic biological processes that are common to very disparate species. It's kind of like the way most fairy tales begin with "Once upon a time". $\endgroup$
    – Barmar
    Commented Jan 22, 2022 at 14:05
  • $\begingroup$ Or the way each chapter of a book begins with "Chapter ##". $\endgroup$
    – Barmar
    Commented Jan 22, 2022 at 14:07

so it might look like: [A,C,C,C,A,T,G] and [A,C,C,C,A,A,T,G], now how can we say what corresponds to what? Do we say there is a SNP at the fifth/sixth base pair?

We say the second one has an insertion at 5.

Have you read the vcf format? It describes how all kinds of variants are documented.


What about something more different like [A,C,C,C,A,A,T,G] and [C,C,A,A]? How do we know what corresponds to what? What are the SNPs?

Absent any context, why would you want to compare two completely different sequences like that?

If you really wanted to compare them, surely single nucleotide polymorphisms isn't the right way to think about the differences.

  • $\begingroup$ I don't yet have a clear picture of what an organism's genome is 'supposed' to look like. Thank you for your answer. $\endgroup$ Commented Jan 21, 2022 at 5:39

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .