VCF files explained: how to open one and what's actually inside
If a whole-genome provider, a clinical lab, or a service like Genotek handed you a file ending in .vcf, you are holding the closest thing genomics has to a universal format. VCF stands for Variant Call Format, and unlike most things in bioinformatics, it is just text. You can open it, read it, and with ten minutes of orientation, actually understand it.
How to open it
Any text editor works: Notepad, TextEdit, VS Code. Big files may make a simple editor sweat, so on large VCFs it is easier to peek at the top with a command line (head -50 yourfile.vcf) or use a viewer that streams the file. If your file ends in .vcf.gz, it is compressed; unzip it first (double-click on most systems, or gunzip yourfile.vcf.gz).
There is no need for special "VCF viewer" software just to look inside. It is text all the way down.
The anatomy
A VCF has two parts. The top is a header: lines starting with ## that describe where the file came from, which reference genome build it uses (GRCh37 or GRCh38, this matters for position numbers), and what the columns mean. Then comes one line per genetic position, with tab-separated columns. The ones that matter to a human:
| Column | Meaning |
|---|---|
| CHROM, POS | Which chromosome, and where on it |
| ID | The variant's name, usually an rsid like rs1801133 |
| REF | The reference genome's letter at this spot |
| ALT | The alternative letter(s) seen here |
| The last column | Your genotype, encoded as numbers |
Reading your genotype
The number code is the part nobody explains. 0 means the REF letter, 1 means the ALT letter, and you have two copies of most positions, so:
0/0= two reference copies. If REF is G and ALT is A, you are G/G.0/1= one of each: G/A.1/1= two alternative copies: A/A.
A real example: the line chr1 11856378 rs1801133 G A ... 0/0 says that at rs1801133, the famous MTHFR position, this person carries G/G. One subtlety worth knowing before you google your letters: the letters in genotype files often differ from the letters used in articles because file and papers can describe opposite DNA strands. That G/G is what papers call C/C.
VCF vs the 23andMe-style file
Consumer chip exports (23andMe, AncestryDNA) list genotypes at a fixed set of a few hundred thousand positions. A VCF can do that too, but VCFs from sequencing can also be enormous, gigabytes for a whole genome, because they cover vastly more positions. The format also carries quality scores and depth information chips do not have. In practice: a chip export tells you "here are the positions we always test", a sequencing VCF tells you "here is what we saw in your actual reads".
What to do with it
Most consumer DNA tools accept chip exports but choke on VCF. Helisoma reads standard VCF directly, in your browser, up to 200 MB of plain .vcf, and turns it into the same wellness report chip users get: replicated findings, honest effect sizes, every study linked. If your whole-genome VCF is larger than that, ask your provider for a sites-only or array-style export, which contains everything a wellness report needs at a fraction of the size.
The short version: your VCF is not a black box. It is a spreadsheet in a trench coat, and the interesting parts are entirely readable.