guavacode

how many times is anything encoded in the guava genome? type a word (or a Konami code) and this streams all 443,755,635 bases of Psidium guajava from NCBI through your browser and counts it, three ways, on both strands: 2-bit text (UTF-8 bytes, A=00 C=01 G=10 T=11), ACGT letters (only if the word is spelled with A, C, G, T), and protein (one-letter amino acids in all six reading frames). a hit is a hit even when it overlaps another.

look up (one per line)

one run downloads about 444 MB from NCBI (all 44 sequences of assembly guava_v11.23) and counts every term in a single pass, so put everything you're curious about in the box first. nothing leaves your browser except the NCBI requests.

or: the first names in your follows

reads the first word of each display name you follow (or the start of their handle) and fills the box above with the distinct names. only names are counted, never who has them. then count, and the top row is the winner.

rank by in

term	follows	2-bit text	ACGT letters	protein

counts add both strands; hover a number for the strand split and what a random genome of the same base mix would give. short words win on raw hits because they're short, and the genome's repeats make real counts differ a lot from chance, which is what "above chance" ranks. text is counted in as-typed, lower, UPPER and Capitalised forms.