🧬 Doing Biology in Julia 🧬

The central dogma of molecular biology

DNA copied into DNA through “replication”
DNA read into RNA through “transcription”
RNA read into protein through “translation”

Major biological molecules are 1D arrays (polymers)

DNA

Protein

The functions of biomolecules are determined by their sequence

Evolutionary relationships are inferred from the relationship between sequences

Important biological algorithms are about alignment and search

BioSequences.jl efficiently encodes biological sequences

import Base: summarysize
using BioSequences, Random

seq = randseq(DNAAlphabet{2}(), 512);
str = String(seq);

summarysize(seq) # 184
summarysize(str) # 520import Base: summarysize
using BioSequences, Random

seq = randseq(DNAAlphabet{2}(), 512);
str = String(seq);

summarysize(seq) # 184
summarysize(str) # 520import Base: summarysize
using BioSequences, Random

seq = randseq(DNAAlphabet{2}(), 512);
str = String(seq);

summarysize(seq) # 184
summarysize(str) # 520

help?> DNAAlphabet{2}
  DNA nucleotide alphabet.

  DNAAlphabet has a parameter N which is a number that determines
  the BitsPerSymbol trait. Currently supported values of N are 2 and 4.

julia> @btime findall(DNA_A, $seq);
  1.487 μs (4 allocations: 1.92 KiB)

julia> @btime findall(==('A'), $str);
  3.093 μs (4 allocations: 1.92 KiB)

BioSequences.jl competes with special-purpose languages

File parsing is a critical part of bioinoformatics

FASTA

> some header | other info
AATTACGC
> foo
AGGGAGATCCC

FASTQ

@ some header | other info
AATTACGC
+
AAAAA#EE
@ foo
AGGGAGATCCC
+
AA#G</EEA6E

SAM

@HD     VN:1.0  SO:unsorted
@SQ     SN:1455__A0A0C2TZA5__A3781_04875        LN:1008
@SQ     SN:1455__A0A0C2XRW0__A3781_18225        LN:804
@SQ     SN:1455__A0A0C2XXQ4__A3781_14565        LN:867
...
VH01194:15:AAAWT2VHV:1:1101:49456:1398:N:0:GAACTGAGCG+CGCTCCACGA#0/1__1.101 16  1134687__A0A378ENW4__cobJ   67  3   150M    *   0   0   CTGCAGGCGGCGGAAATCGTCGTCGGTTATAAAACTTACACCCATCTGGTGAAGGCTTTTACCGGCGACAAGCAGGTGATCAAAACCGGGATGTGCAAAGAGATTGAACGCTGTCAGGCGGCGATTGAACTGGCGCAGGCCGGGCACAAC  CCCCCCCCCCCCCCCC;CCCCCCCCCCCC;CCCCCC;CCCCCCC;CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC  AS:i:-59    XN:i:0  XM:i:12 XO:i:0  XG:i:0  NM:i:12 MD:Z:23T2C8C5G2C8A2C11T38G5C14G17T3 YT:Z:UU
VH01194:15:AAAWT2VHV:1:1101:31259:2098:N:0:GAACTGAGCG+CGCTCCACGA#0/1__1.441 16  73098__A0A1B7K6J8__M989_00754   328 40  150M    *   0   0   CCAGCCGATTTCAGGAAATTAGGCCGTGATGCCGCGGCGACGCTGTTGTCGGTATCTAACGTAACGCTCTGGAATTCCATCGACTATTTCAGCCCCAGCGCCGAGCATAATCCTTTATTGATGACCTGGTCATTGGGCGTGGAAGAACAG  CC-CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC;CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC;CCCCCCCCCCCCCCCCCCCCCCCCCCCCCCCC;CCCCCCCCCCCCCC;CCCCCCCCCCCCCCCCC  AS:i:-18    XN:i:0  XM:i:4  XO:i:0  XG:i:0  NM:i:4  MD:Z:2C23C18C16T87  YT:Z:UU

The development of every bioinformatics tool begins with the definition of a new file format, incompatible with all previous formats.

- Charles Darwin

Automa.jl enables the construction of correct but efficient file parsers

Automa makes Deterministic Finite Automata

fasta_regex = let
    header = re"[a-z]+"
    seqline = re"[ACGT]+"
    record = '>' * header * '\n' * rep1(seqline * '\n')
    rep(record)
end

machine = let
    header = onexit!(onenter!(re"[a-z]+", :mark_pos), :header)
    seqline = onexit!(onenter!(re"[ACGT]+", :mark_pos), :seqline)
    record = onexit!(re">" * header * '\n' * rep1(seqline * '\n'), :record)
    compile(rep(record))
end

BioMakie.jl enables easy plotting of protein structure

using BioMakie
using GLMakie
using BioStructures
struc = retrievepdb("2vb1") |> Observable
## or
struc = read("2vb1.pdb", BioStructures.PDB) |> Observable

fig = Figure()
plotstruc!(fig, struc; plottype = :ballandstick, gridposition = (1,1), atomcolors = aquacolors)
plotstruc!(fig, struc; plottype = :covalent, gridposition = (1,2))