GitHub

Melting Temperature Calculation

SeqFold.jl provides functions for calculating DNA melting temperatures using nearest-neighbor thermodynamics.

Showcase

julia> seq = "GGGAGGTCAGCAAACCTGAACCTGTTGAGATGTTGACGTCAGGAAACCCT";

julia> tm(seq) # melting at PCR conditions, same as tm(seq, conditions=:pcr)
80.4

julia> MeltingConditions(:pcr) # here they are as a preset inherited from seqfold library
MeltingConditions (PCR preset)
  • NEB PCR buffer conditions for Taq DNA Polymerase
  • seq1 concentration: 250.0 nM (typical primer concentration)
  • seq2 concentration: 0.0 nM (asymmetric PCR)
  • Mg²⁺: 1.5 mM (optimal for Taq DNA Polymerase per NEB guidelines)
  • K⁺: 50.0 mM
  • Tris: 2.0 mM
  • dNTPs: 0.2 mM

julia> tm(seq, Mg=5) # altering default conditions
81.4

julia> MeltingConditions(:std) # second preset, inherited from seqfold library
MeltingConditions (standard preset)
  • Standard hybridization buffer
  • seq1 concentration: 25.0 nM
  • seq2 concentration: 25.0 nM
  • Na⁺: 50.0 mM

julia> tm(seq, conditions=:std, Na=150) # altering chosen preset
79.8

julia> custom_conds = MeltingConditions( # Specifying custom conditions
       seq1_conc=30,seq2_conc=20,Na=5,K=5,Tris=15,Mg=0,dNTPs=0)
MeltingConditions (custom)
  • seq1 concentration: 30.0 nM
  • seq2 concentration: 20.0 nM
  • Na⁺: 5.0 mM
  • K⁺: 5.0 mM
  • Tris: 15.0 mM

julia> tm(seq, conditions=custom_conds, Tris=50) # flexible adjustments
68.4

Exported names

SeqFold.tm — Function
tm(seq1, seq2; conditions=:pcr, kwargs...) -> Float64
tm(seq; conditions=:pcr, kwargs...)  -> Float64

Calculate melting temperature (Tm °C) for DNA duplex formation using nearest-neighbor thermodynamics.

Arguments

  • seq1::AbstractString: DNA sequence;
  • seq2::AbstractString: Another DNA sequence, must be the same length as seq1;
  • seq::AbstractString: Single DNA sequence to be matched with its exact complement;
  • conditions: Buffer conditions specification (see below);
  • kwargs...: Additional parameters to override preset conditions in place.

Conditions Specification

The conditions parameter can be:

  • :pcr (default) or :std: Use preset conditions;
  • MeltingConditions object: Custom conditions (for more info see: MeltingConditions);
  • NTuple{7, Float64}: (seq1_conc, seq2_conc, Na, K, Tris, Mg, dNTPs);
  • NamedTuple with condition fields.

Examples

julia> tm("GGGGGG")
15.5

julia> tm("GGGGGG", conditions=:pcr)
15.5

julia> tm("GGGGGG", Mg=10)
23.0

julia> tm("GGGGGG", conditions=:std)
2.6

julia> tm("GGGGGG", conditions=:std, Na=100)
5.1

julia> tm("GGGGGG", conditions=MeltingConditions(150, 150, 20, 0, 0, 10, 0))
16.8

julia> tm("ACCCCC", "GGGGGG")
6.2

Implementation

The calculation uses nearest-neighbor thermodynamic parameters from related literature (follow links below to see sources), accounting for initialization terms, nearest-neighbor pairs, and terminal mismatches when present.

See also

MeltingConditions

source
SeqFold.MeltingConditions — Type
MeltingConditions(seq1_conc, seq2_conc, Na, K, Tris, Mg, dNTPs)

Represents buffer conditions for melting temperature calculation.

Fields and Units

  • seq1_conc, seq2_conc: Sequence concentrations in nM (nanomolar)
  • Na, K, Tris, Mg, dNTPs: Buffer component concentrations in mM (millimolar)

Examples

julia> MeltingConditions(:pcr)
MeltingConditions (PCR preset)
  • NEB PCR buffer conditions for Taq DNA Polymerase
  • seq1 concentration: 250.0 nM (typical primer concentration)
  • seq2 concentration: 0.0 nM (asymmetric PCR)
  • Mg²⁺: 1.5 mM (optimal for Taq DNA Polymerase per NEB guidelines)
  • K⁺: 50.0 mM
  • Tris: 2.0 mM
  • dNTPs: 0.2 mM

julia> MeltingConditions(:std)
MeltingConditions (standard preset)
  • Standard hybridization buffer
  • seq1 concentration: 25.0 nM
  • seq2 concentration: 25.0 nM
  • Na⁺: 50.0 mM

julia> MeltingConditions(seq1_conc=10,seq2_conc=0,Na=0,K=0,Tris=10,Mg=0,dNTPs=0)
MeltingConditions (custom)
  • seq1 concentration: 10.0 nM
  • seq2 concentration: 0.0 nM
  • Tris: 10.0 mM

julia> MeltingConditions(10,0,0,0,0,5,0)
MeltingConditions (custom)
  • seq1 concentration: 10.0 nM
  • seq2 concentration: 0.0 nM
  • Mg²⁺: 5.0 mM (higher than NEB's recommended 1.5-2.0 mM range)

julia> MeltingConditions(seq1_conc=0,seq2_conc=0,Na=5,K=0,Tris=0,Mg=0,dNTPs=0)
ERROR: ArgumentError: DNA concentration is too low!
[...]

julia> MeltingConditions(seq1_conc=10,seq2_conc=10,Na=0,K=0,Tris=0,Mg=0,dNTPs=0)
ERROR: ArgumentError: No cations for salt correction (Mg=0, Na=0, K=0, Tris=0)
[...]

See also

tm, SeqFold.tm_cache

source

Public names

SeqFold.tm_cache — Function
SeqFold.tm_cache(seq1, seq2; conditions=:pcr, kwargs...) -> Matrix{Float64}
SeqFold.tm_cache(seq; conditions=:pcr, kwargs...) -> Matrix{Float64}

Compute a matrix of melting temperatures for all possible subsequences of a DNA sequence pair.

Arguments

  • seq1::AbstractString: DNA sequence;
  • seq2::AbstractString: Another DNA sequence, must be the same length as seq1;
  • seq::AbstractString: Single DNA sequence to be matched with its exact complement;
  • conditions: Buffer conditions specification (see tm docstrings);
  • kwargs...: Additional parameters to override preset conditions in place.

Returns

A Matrix{Float64} where element [i, j] contains the melting temperature (in °C) of the subsequence from position i to position j, inclusive. Elements where j < i contain NaN as they represent invalid ranges, and single-nucleotide subsequences also have NaN as they don't have meaningful Tm values.

Examples

julia> SeqFold.tm_cache("ATCC")
4×4 Matrix{Float64}:
 NaN  -212.6   -95.3   -48.6
 NaN   NaN    -161.6   -82.7
 NaN   NaN     NaN    -135.5
 NaN   NaN     NaN     NaN

julia> SeqFold.tm_cache("AAGC", "TTCG")
4×4 Matrix{Float64}:
 NaN  -204.8   -94.6   -40.3
 NaN   NaN    -166.7   -72.9
 NaN   NaN     NaN    -116.7
 NaN   NaN     NaN     NaN

julia> SeqFold.tm_cache("AAGC", "TTCG"; conditions=:std)
4×4 Matrix{Float64}:
 NaN  -213.1  -109.3   -55.6
 NaN   NaN    -177.9   -88.1
 NaN   NaN     NaN    -129.8
 NaN   NaN     NaN     NaN

Implementation

The function uses dynamic programming to build a cache of Tm values for all subsequences. The algorithm has O(n²) time and space complexity, where n is the sequence length. This approach avoids redundant calculations when multiple Tm values for different subsequences are needed.

See also

tm, SeqFold.gc_cache, SeqFold.dg_cache

source