#Composition and properties

Runnable: jennifer run examples/composition.j. The parameter sets behind these numbers are exported; see Parameter constants.

#Composition

bioseq.baseCount("AATTGGCC");       # {"A": 2, "T": 2, "G": 2, "C": 2}
bioseq.baseFrequency($gene);        # the same, as percentages
bioseq.residueCount($protein);
bioseq.residueFrequency($protein);

Only residues present get an entry; guard a lookup with maps.has.

#GC content

bioseq.gcContent("ATGCGC");   # 66.67
bioseq.atContent("ATGCGC");   # 33.33
bioseq.gcWindows($genome, 100, 50);   # list of float, one per window

Ambiguity codes count toward the length but toward neither G/C nor A/T/U, so on a sequence with N (or S) the two sum to less than 100.

gcWindows(s, size, step) starts a window every step bases and drops the last partial window, so every value covers exactly size bases; windows cuts identically. A sequence shorter than one window raises.

#Molecular weight

bioseq.molecularWeight("ATGC", bioseq.DNA);       # 1173.83
bioseq.molecularWeight($rna, bioseq.RNA);
bioseq.molecularWeight($protein, bioseq.PROTEIN);

Average masses, in daltons.

• DNA, RNA: one strand with a free 5' hydroxyl (an unphosphorylated synthetic oligo): residue masses + WATER_AVG_MASS - PHOSPHATE_MASS. Roughly double it for a duplex.
• PROTEIN: residue masses (avgMass of each AminoAcid)
  • one water, the ExPASy convention.

Ambiguity codes raise (why).

#Melting temperature

bioseq.meltingTemp("ATGCATGCATGCATGC");                     # 48.0, Wallace rule
bioseq.meltingTempNn("ATGCATGCATGCATGC");                   # 50.7, 250 nM primer, 50 mM Na+
bioseq.meltingTempNnWith("ATGCATGCATGCATGC", 250.0, 100.0); # 53.9, 100 mM Na+

Both take DNA only, in degrees Celsius, and raise on ambiguity codes.

Wallace rule: 2(A+T) + 4(G+C). Calibrated for oligos of about 14 to 20 bases; on longer sequences it returns a number with no physical meaning.

Nearest-neighbour: SantaLucia (1998) unified parameters (NN_ENTHALPY, NN_ENTROPY, GAS_CONSTANT). Enthalpy and entropy are summed over the dinucleotides, with helix initiation per terminal base, a salt correction and a self-complementarity correction, then Tm = dH / (dS + R ln(Ct/4)) - 273.15. Needs at least two bases; concentrations must be positive (primerNm in nM, saltMm in mM). Other parameter sets differ by a degree or two.

Order matters to the stacking model, not to Wallace:

#Protein chemistry

bioseq.netCharge($protein, 7.0);
bioseq.isoelectricPoint($protein);
bioseq.gravy($protein);
bioseq.aromaticity($protein);

All four raise on an empty or non-protein sequence; the first three also on B, Z, X, J or *.

Net charge sums the fractional charge of the two termini and the side chains of D, E, C, Y (acidic) and H, K, R (basic). It decreases monotonically with pH.

Isoelectric point: bisection on the net charge over pH 0 to 14, stopping within 1e-4 of zero or after 100 steps. The pKa values are the EMBOSS iep set (PKA_N_TERMINUS, PKA_C_TERMINUS, AminoAcid.pKa); EMBOSS, ExPASy and Sillero differ by a few tenths of a unit, so compare against iep, not ProtParam.

GRAVY: mean Kyte-Doolittle hydropathy. Positive is hydrophobic; soluble globular proteins typically score -0.5 to -1.5. The preproinsulin signal fragment in testdata/ scores +1.25.

Aromaticity: fraction of F, W and Y, 0 to 1.

#Summaries

def s as bioseq.DnaStats init bioseq.dnaStats("ATGCNNNN");
$s.gc;               # 25.0
$s.ambiguous;        # true
$s.molecularWeight;  # 0.0
def p as bioseq.ProteinStats init bioseq.proteinStats($protein);

One call per report row. With ambiguity codes the summary still has length, counts and GC/AT (DNA) or aromaticity (protein); mass, Tm, pI and GRAVY are 0.0 and ambiguous is true. dnaStats rejects RNA. DnaStats.meltingTemp is the Wallace figure. Fields: cheatsheet.

#Codon usage

for (def u in bioseq.codonUsage($cds)) {
    io.printf("%s %s %d %f|prec=2\n", $u.codon, $u.aminoAcid, $u.count, $u.fraction);
}

One CodonUse per codon present, in first-seen order. fraction is the share among codons for the same residue (each residue's fractions sum to 1.0); perThousand sums to 1000. In the insulin fragment CTG is 6 of the 8 leucines (0.75). Codons are reported DNA-spelled; a codon the table does not hold (one with an ambiguity code) counts under residue X. codonUsageWith takes a codon table.