BioCollections

   

Summary: This example shows how to use collections of biological items in BioFSharp

Analogous to the build-in collections BioFSharp provides BioSeq, BioList and BioArray for individual collection specific optimized operations. The easiest way to create them are the ofBioItemString -functions

open BioFSharp

let s1 = "PEPTIDE" |> BioSeq.ofAminoAcidString 
let s2 = "PEPTIDE" |> BioList.ofAminoAcidString 
let s3 = "TAGCAT"  |> BioArray.ofNucleotideString 

//Peptide represented as a Bioseq
"PEPTIDE" |> BioSeq.ofAminoAcidString 

Warning: Output, it-value and value references require --eval

//Peptide represented as a BioList
"PEPTIDE"|> BioList.ofAminoAcidString 

Warning: Output, it-value and value references require --eval

//Nucleotide sequence represented as a BioArray
"TAGCAT" |> BioArray.ofNucleotideString 

Warning: Output, it-value and value references require --eval

Nucleotides

Figure 1: Selection of covered nucleotide operations (A) Biological principle. (B) Workflow with BioSeq. (C) Other covered functionalities.

Let's imagine you have a given gene sequence and want to find out what the according protein might look like.

let myGene = BioArray.ofNucleotideString "ATGGCTAGATCGATCGATCGGCTAACGTAA"

Warning: Output, it-value and value references require --eval

Yikes! Unfortunately we got the 5'-3' coding strand. For proper transcription we should get the complementary strand first:

let myProperGene = BioArray.complement myGene

Warning: Output, it-value and value references require --eval

Now let's transcribe and translate it:

let myTranslatedGene = 
    myProperGene
    |> BioArray.transcribeTemplateStrand
    |> BioArray.translate 0

Warning: Output, it-value and value references require --eval

Of course, if your input sequence originates from the coding strand, you can directly transcribe it to mRNA since the only difference between the coding strand and the mRNA is the replacement of 'T' by 'U' (Figure 1B)

let myTranslatedGeneFromCodingStrand = 
    myGene
    |> BioArray.transcribeCodingStrand
    |> BioArray.translate 0

Warning: Output, it-value and value references require --eval

Other Nucleotide conversion operations are also covered:

let mySmallGene = BioSeq.ofNucleotideString  "ATGTTCCGAT"

Warning: Output, it-value and value references require --eval

let smallGeneRev = BioSeq.reverse mySmallGene 

Warning: Output, it-value and value references require --eval

let smallGeneComp    = BioSeq.complement mySmallGene

Warning: Output, it-value and value references require --eval

let smallGeneRevComp = BioSeq.reverseComplement mySmallGene

Warning: Output, it-value and value references require --eval

AminoAcids

Basics

Some functions which might be needed regularly are defined to work with nucleotides and amino acids:

let myPeptide = "PEPTIDE" |> BioSeq.ofAminoAcidString 

Warning: Output, it-value and value references require --eval

let myPeptideFormula = BioSeq.toFormula myPeptide |> Formula.toString 

Warning: Output, it-value and value references require --eval

let myPeptideMass = BioSeq.toAverageMass myPeptide 

Warning: Output, it-value and value references require --eval

Digestion

BioFSharp also comes equipped with a set of tools aimed at cutting apart amino acid sequences. To demonstrate the usage, we'll throw some trypsin at the small RuBisCO subunit of Arabidopos thaliana:
In the first step, we define our input sequence and the protease we want to use.

let RBCS = 
    """MASSMLSSATMVASPAQATMVAPFNGLKSSAAFPATRKANNDITSITSNGGRVNCMQVWP
    PIGKKKFETLSYLPDLTDSELAKEVDYLIRNKWIPCVEFELEHGFVYREHGNSPGYYDGR
    YWTMWKLPLFGCTDSAQVLKEVEECKKEYPNAFIRIIGFDNTRQVQCISFIAYKPPSFT""" 
    |> BioArray.ofAminoAcidString

Warning: Output, it-value and value references require --eval

let trypsin = Digestion.Table.getProteaseBy "Trypsin"

With these two things done, digesting the protein is a piece of cake. For doing this, just use the digest function.

let digestedRBCS = Digestion.BioArray.digest trypsin 0 RBCS 

digestedRBCS
|> Seq.head

Warning: Output, it-value and value references require --eval

(*
In reality, proteases don't always completely cut the protein down. Instead, some sites stay intact and should be considered for in silico analysis. 
This can easily be done with the `concernMissCleavages` function. It takes the minimum and maximum amount of misscleavages you want to have and also the digested protein. As a result you get all possible combinations arising from this information.
*)

let digestedRBCS' = Digestion.BioArray.concernMissCleavages 0 2 digestedRBCS

digestedRBCS
|> Seq.item 1

Warning: Output, it-value and value references require --eval