tm_nn: Calculate melting temperature using nearest neighbor...

In TmCalculator: Genome-Wide Nucleic Acid Melting Temperature Profiling and Multi-Omics Integration

Calculate melting temperature using nearest neighbor thermodynamics

Description

Calculates melting temperature (Tm) from nearest-neighbor (NN) thermodynamics, summing the stacking enthalpies and entropies of adjacent base-pair steps and applying initiation, symmetry, salt and chemical corrections. Terminal mismatches, internal mismatches and dangling ends are supported through dedicated parameter tables. The function verifies that every dinucleotide step in the input sequence is present in the selected tables before calculating.

Usage

tm_nn(
  gr_seq,
  ambiguous = FALSE,
  shift = 0,
  nn_table = c("DNA_NN_SantaLucia_2004", "DNA_NN_Ghosh_2020_PEG200",
    "DNA_NN_Breslauer_1986", "DNA_NN_Sugimoto_1996", "DNA_NN_Allawi_1998",
    "RNA_NN_Freier_1986", "RNA_NN_Xia_1998", "RNA_NN_Chen_2012", "RNA_NN_Zuber_2022",
    "RNA_NN_Ghosh_2023_PEG200", "RNA_DNA_NN_Sugimoto_1995", "DNA_NN_Weber_2015",
    "DNA_NN_Weber_OW04_69", "DNA_NN_Weber_OW04_119", "DNA_NN_Weber_OW04_220",
    "DNA_NN_Weber_OW04_621", "DNA_NN_Weber_OW04_1020", "RNA_NN_Weber_VIF_71",
    "RNA_NN_Weber_VIF_121", "RNA_NN_Weber_VIF_221", "RNA_NN_Weber_VIF_621", 
    
    "RNA_NN_Weber_VIF_1021", "RNA_NN_Weber_FIF_71", "RNA_NN_Weber_FIF_121",
    "RNA_NN_Weber_FIF_221", "RNA_NN_Weber_FIF_621", "RNA_NN_Weber_FIF_1021",
    "RNA_DNA_NN_Weber_2019_FT", "RNA_DNA_NN_Weber_2019_VH", "RNA_DNA_NN_Weber_2019_LS",
    "RNA_DNA_NN_Banerjee_2020"),
  tmm_table = "DNA_TMM_Bommarito_2000",
  imm_table = "DNA_IMM_Peyret_1999",
  de_table = c("DNA_DE_Bommarito_2000", "RNA_DE_Turner_2010"),
  dnac_high = 25,
  dnac_low = 25,
  self_comp = FALSE,
  Na = 50,
  K = 0,
  Tris = 0,
  Mg = 0,
  dNTPs = 0,
  salt_method = c("Schildkraut2010", "Wetmur1991", "SantaLucia1996", "SantaLucia1998-1",
    "SantaLucia1998-2", "Owczarzy2004", "Owczarzy2008", "none"),
  DMSO = 0,
  formamide_unit = list(value = 0, unit = "percent"),
  dmso_factor = 0.75,
  formamide_factor = 0.65
)

Arguments

gr_seq	Sequence(s) in 5' to 3' direction, as the GRanges that to_genomic_ranges returns. A character vector of sequences, a path to a FASTA file, or genomic coordinate strings are also accepted and converted for you. To pair a sequence with a complement of your own, build the object explicitly: to_genomic_ranges(seq, complement_seq = cmp).
ambiguous	Logical value controlling how ambiguous bases are handled: - TRUE: Ambiguous bases (e.g., N, R, Y) are included in calculations - FALSE (default): Ambiguous bases are excluded from calculations
shift	Integer value controlling the alignment offset between primer and template sequences. Visual representation of different shift values: shift = 0 (default): Primer: 5' ATGCG 3' Template: 3' TACGC 5' shift = -1: Primer: 5' ATGCG 3' Template: 3' TACGC 5' ^ shift = 1: Primer: 5' ATGCG 3' Template: 3' TACGC 5' ^ The shift parameter is necessary when: - Sequences have different lengths - Dangling ends are required - Specific alignment positions are needed
nn_table	Thermodynamic nearest-neighbor parameters for different nucleic acid hybridizations. Parameter sets are listed below by hybridization type. Sets marked with a sodium concentration were fitted at that condition and are not salt-corrected further (see the "Choosing a parameter set" section). DNA/DNA hybridizations, reference salt: - "DNA_NN_Breslauer_1986": Original DNA/DNA parameters - "DNA_NN_Sugimoto_1996": Improved DNA/DNA parameters - "DNA_NN_Allawi_1998": Watson-Crick parameters from Allawi & SantaLucia (1997), Table 1; the historical identifier is retained for compatibility - "DNA_NN_SantaLucia_2004": Unified DNA/DNA parameters (default) DNA/DNA hybridizations, melting-temperature optimized (Weber 2015): - "DNA_NN_Weber_2015": Combined dataset, 1020 mM. Recommended when a salt-optimized DNA set is wanted at high salt - "DNA_NN_Weber_OW04_69", "...119", "...220", "...621", "...1020": fitted independently at 69, 119, 220, 621 and 1020 mM sodium DNA/DNA under molecular crowding (cell-like rather than dilute solution): - "DNA_NN_Ghosh_2020_PEG200": fitted in 40 wt RNA/RNA hybridizations, reference salt: - "RNA_NN_Freier_1986": Original RNA/RNA parameters - "RNA_NN_Xia_1998": Improved RNA/RNA parameters - "RNA_NN_Chen_2012": Updated RNA/RNA parameters with GU pair corrections - "RNA_NN_Zuber_2022": Successor to Xia 1998 with improved end effects. The terminal-AU penalty is replaced by end terms that depend on the penultimate base pair, applied automatically from a companion table RNA/RNA under molecular crowding (cell-like rather than dilute solution): - "RNA_NN_Ghosh_2023_PEG200": fitted in 40 wt and shown to describe duplexes in an intracellular cation composition RNA/RNA hybridizations, salt-optimized (Ferreira 2019). VIF (variable initiation factors) gave better cross-validation than FIF (fixed): - "RNA_NN_Weber_VIF_71", "...121", "...221", "...621", "...1021" - "RNA_NN_Weber_FIF_71", "...121", "...221", "...621", "...1021" RNA/DNA hybridizations: - "RNA_DNA_NN_Sugimoto_1995": RNA/DNA hybridization parameters - "RNA_DNA_NN_Weber_2019_FT": curve-fitting derived, 1000 mM. Best performing high-salt hybrid set in Basilio Barbosa (2019) - "RNA_DNA_NN_Weber_2019_VH": van't Hoff derived, 1000 mM - "RNA_DNA_NN_Weber_2019_LS": low salt, 100 mM - "RNA_DNA_NN_Banerjee_2020": improved hybrid parameters fitted at a physiological condition (100 mM NaCl), Banerjee et al. (2020) For every hybrid set the sequence you supply must be the RNA strand, written 5' to 3' and spelled with T in place of U; its complement is then the DNA strand, read 3' to 5'. The published keys are indexed the same way, RNA on top: "AG/TC" means 5'-rAG-3' paired with 3'-dTC-5'. Unlike DNA/DNA and RNA/RNA parameters, hybrid parameters are not invariant under reversing a key, because reversing it swaps which strand carries the ribose. Supplying the DNA strand therefore does not give the same answer: it silently reports the melting temperature of the opposite hybrid. The asymmetry is large, since a purine-rich RNA strand binds DNA more tightly than a pyrimidine-rich one (for example 5'-rAA-3'/3'-dTT-5' and 5'-rUU-3'/3'-dAA-5' differ by about 0.8 kcal/mol in delta G at 37 degrees C). Alternatively, supply a matrix or data.frame of parameters directly. This is the route for parameter sets the package does not ship, in particular sets covering modified bases such as 5-methylcytosine. Requirements: numeric, with columns 1 and 2 read as delta H (kcal/mol) and delta S (cal/mol/K); further columns are ignored; row names giving the parameter keys, e.g. "AA/TT", "init", "init_A/T", "sym"; every key of a built-in reference set must be present. The reference is named by attr(x, "reference"), or defaults to the first built-in listed for the argument, which is a DNA/DNA set; RNA and hybrid tables should therefore set the attribute. Extra keys beyond the reference are kept, which is how a modified-base set adds stacks rather than replacing them. The supplied table is reordered to the reference key order before use, so that two tables differing only in row order give identical results. A missing key would otherwise contribute zero to the calculation instead of raising an error, which is why the full key set is required. Keys that disagree with their reverse complement produce a warning: expected for modified bases, a transposition error otherwise. Two optional attributes are honoured. attr(x, "salt_mM") marks a set as fitted at a stated sodium concentration, which suppresses the salt correction at that concentration exactly as for the built-in sets fitted this way; without it the table is treated as a reference-condition set and salt_method is applied. attr(x, "end_table") supplies a companion penultimate-pair end-effect table.
tmm_table	Thermodynamic parameters for terminal mismatches. Default: "DNA_TMM_Bommarito_2000" These 48 parameters come from SantaLucia & Peyret (2001), patent WO2001094611A2, Tables 2-3. The historical identifier is retained; Bommarito (2000) is the source of DNA dangling ends, not this table.
imm_table	Thermodynamic parameters for internal mismatches. Default: "DNA_IMM_Peyret_1999" This is a composite of six publications (1997-2005), not a single Peyret (1999) table: 11 G.T, 8 G.A, 8 A.C, 8 C.T, 16 like-with-like mismatches and 36 inosine entries. The A.C parameters are for pH 7.
de_table	Thermodynamic parameters for dangling ends. Default: "DNA_DE_Bommarito_2000" Available options: - "DNA_DE_Bommarito_2000": Parameters for DNA dangling ends - "RNA_DE_Turner_2010": NNDB Turner 2004 RNA dangling-end compilation; 2010 is the database publication year. Entropies are derived from tabulated enthalpies and free energies at 37 degrees C
dnac_high	Concentration of the higher concentrated strand in nM. Default: 25 Typically this is the primer (for PCR) or the probe concentration.
dnac_low	Concentration of the lower concentrated strand in nM. Default: 25 This is typically the template concentration.
self_comp	Logical value indicating if the sequence is self-complementary: - TRUE: Sequence can bind to itself, dnac_low is ignored - FALSE (default): Sequence binds to a different complementary sequence
Na	Millimolar concentration of sodium ions. Default: 50
K	Millimolar concentration of potassium ions. Default: 0
Tris	Millimolar concentration of Tris buffer. Default: 0
Mg	Millimolar concentration of magnesium ions. Default: 0
dNTPs	Millimolar concentration of deoxynucleotide triphosphates. Default: 0
salt_method	Salt correction method. Options are: Available options: - "Schildkraut2010": Schildkraut & Lifson (1965); historical identifier - "Wetmur1991": Classic salt correction method - "SantaLucia1996": DNA-specific salt correction - "SantaLucia1998-1": Improved DNA salt correction, applied to Tm - "SantaLucia1998-2": the same correction applied to the entropy of the nearest-neighbor model rather than to Tm, which is why it is available here and not in tm_gc - "Owczarzy2004": Comprehensive salt correction - "Owczarzy2008": Updated comprehensive salt correction - "none": Disables salt correction entirely Note: Parameter sets fitted at a specific sodium concentration (those carrying a "salt_mM" attribute, e.g. the Weber/VarGibbs series and Banerjee 2020) already account for salt. When the requested Na matches the concentration such a set was fitted at, salt correction is skipped automatically; when it does not, a warning is issued.
DMSO	Percent DMSO concentration in the reaction mixture. Default: 0 DMSO can lower the melting temperature of nucleic acid duplexes.
formamide_unit	Formamide concentration as 'list(value, unit)'. Default: list(value = 0, unit = "percent") - value: numeric value of formamide concentration - unit: character string specifying the unit ("percent" or "molar") Default: list(value=0, unit="percent")
dmso_factor	Coefficient of melting temperature (Tm) decrease per percent DMSO. Default: 0.75 (von Ahsen N, 2001, PMID:11673362) Other accepted empirical coefficients: 0.5, 0.6, 0.65, 0.675
formamide_factor	Coefficient of melting temperature (Tm) decrease per percent formamide. Default: 0.65, an empirical convention. Accepted alternatives are 0.6 and 0.72; these are not all estimates from the same publication. The molar formula instead follows Blake & Delcourt (1996).

Details

DNA_NN_Breslauer_1986: Breslauer K J (1986) <doi:10.1073/pnas.83.11.3746>

DNA_NN_Sugimoto_1996: Sugimoto N (1996) <doi:10.1093/nar/24.22.4501>

DNA_NN_Allawi_1998: Allawi H T & SantaLucia J Jr (1997), Table 1 <doi:10.1021/bi962590c>; also SantaLucia (1998), Table 2 <doi:10.1073/pnas.95.4.1460>

DNA_NN_SantaLucia_2004: SantaLucia J (2004) <doi:10.1146/annurev.biophys.32.110601.141800>

RNA_NN_Freier_1986: Freier S (1986) <doi:10.1073/pnas.83.24.9373>

RNA_NN_Xia_1998: Xia T (1998) <doi:10.1021/bi9809425>

RNA_NN_Chen_2012: Chen JL (2012) <doi:10.1021/bi3002709>

RNA_DNA_NN_Sugimoto_1995: Sugimoto N (1995)<doi:10.1021/bi00035a029>

The following sets were derived by melting-temperature optimization and are fitted at the sodium concentration given in parentheses. They are not salt-corrected further; see the “Choosing a parameter set” section.

DNA_NN_Weber_2015 (1020 mM): Weber G (2015) <doi:10.1093/bioinformatics/btu751>

DNA_NN_Weber_OW04_69 (69 mM), DNA_NN_Weber_OW04_119 (119 mM), DNA_NN_Weber_OW04_220 (220 mM), DNA_NN_Weber_OW04_621 (621 mM), DNA_NN_Weber_OW04_1020 (1020 mM): Weber G (2015) <doi:10.1093/bioinformatics/btu751>

RNA_NN_Weber_VIF_71 (71 mM), RNA_NN_Weber_VIF_121 (121 mM), RNA_NN_Weber_VIF_221 (221 mM), RNA_NN_Weber_VIF_621 (621 mM), RNA_NN_Weber_VIF_1021 (1021 mM): Ferreira I (2019) <doi:10.1016/j.chemphys.2019.01.016>, variable initiation factors

RNA_NN_Weber_FIF_71 (71 mM), RNA_NN_Weber_FIF_121 (121 mM), RNA_NN_Weber_FIF_221 (221 mM), RNA_NN_Weber_FIF_621 (621 mM), RNA_NN_Weber_FIF_1021 (1021 mM): Ferreira I (2019) <doi:10.1016/j.chemphys.2019.01.016>, fixed initiation factors

RNA_DNA_NN_Weber_2019_FT (1000 mM), RNA_DNA_NN_Weber_2019_VH (1000 mM), RNA_DNA_NN_Weber_2019_LS (100 mM): Basilio Barbosa V (2019) <doi:10.1016/j.bpc.2019.106189>

RNA_DNA_NN_Banerjee_2020 (100 mM): Banerjee D (2020) <doi:10.1093/nar/gkaa572>

RNA_NN_Zuber_2022: Zuber J (2022) <doi:10.1093/nar/gkac261>

RNA_NN_Ghosh_2023_PEG200 (100 mM, 40 wt

DNA_NN_Ghosh_2020_PEG200 (100 mM, 40 wt

DNA_TMM_Bommarito_2000: SantaLucia J Jr & Peyret N (2001), WO2001094611A2, Tables 2-3 (https://patents.google.com/patent/WO2001094611A2/en)

DNA_IMM_Peyret_1999: Allawi H T & SantaLucia J Jr (1997, G.T) <doi:10.1021/bi962590c>; (1998, G.A) <doi:10.1021/bi9724873>; (1998, A.C) <doi:10.1021/bi9803729>; (1998, C.T) <doi:10.1093/nar/26.11.2694>; Peyret N et al. (1999, A.A/C.C/G.G/T.T) <doi:10.1021/bi9825091>; Watkins N E Jr & SantaLucia J Jr (2005, inosine) <doi:10.1093/nar/gki918>

DNA_DE_Bommarito_2000: Bommarito S (2000) <doi:10.1093/nar/28.9.1929>

RNA_DE_Turner_2010: NNDB Turner 2004 dangling-end compilation; Turner D H & Mathews D H (2010, database description) <doi:10.1093/nar/gkp892>

Value

A TmCalculator list with:

gr	The input GRanges with metadata columns Tm and GC (melting temperature in ^{\circ}C and GC percent). Sequences that could not be evaluated receive NA.
options	The calculation parameters actually used, including the parameter tables and their citations, ion and additive concentrations, Salt correction method, and two fields recording how salt was handled: Salt correction applied (logical) and Parameter set fitted at [Na+] (mM), which is NA for reference-salt sets. Regions skipped for containing N are listed in Skipped regions containing N.

Choosing a parameter set

Parameter sets fall into two families that are handled differently.

The two families are distinguished by one mechanical criterion: a set is condition-specific if and only if it carries a salt_mM attribute.

Reference-salt sets (no salt_mM; Breslauer 1986, Sugimoto 1996, Allawi 1997 (legacy name: Allawi_1998), SantaLucia 2004, Freier 1986, Xia 1998, Chen 2012, Zuber 2022, Sugimoto 1995) were fitted at a single reference sodium concentration, and other conditions are reached by applying one of the salt_method correction formulas.

Condition-specific sets (the Weber/VarGibbs series, Banerjee 2020, and the molecular-crowding sets of Ghosh 2020 and Ghosh 2023) were instead fitted directly at a stated sodium concentration and are intended to replace salt correction rather than be corrected. When the requested Na matches the set's salt_mM value, salt correction is skipped automatically; when it does not, the correction is still applied but a warning is issued, since correcting an already condition-specific set double-counts the ionic effect. Whether a correction was applied is recorded in the returned options.

As a rule of thumb, pick the set whose fitted salt is closest to your experimental condition rather than correcting a distant one.

What the initiation terms are charged on

A dangling end or a terminal mismatch is accounted for by its own parameter, from de_table or tmm_table, and the position it covers is then consumed. The initiation terms – init, init_A/T, init_G/C, init_5T/A and the init_allA/T / init_oneG/C pair – are charged on what remains, that is, on the duplex that actually closes.

This matters because the terminal AT penalty is a property of the closing base pair. SantaLucia and Hicks (2004) define it as applied for each end of a duplex that has a terminal AT, so the question it asks is whether the pair holding that end shut is AT or GC. A mismatched terminus is neither an AT nor a GC pair, and a dangling residue is not in a base pair at all, so neither one can carry the penalty. Charging it there would also double-count the end, since the terminal-mismatch and dangling-end parameters were measured as terminal and already carry that end's contribution.

Some implementations index these terms on the sequence as supplied rather than on the trimmed duplex. That makes the melting temperature depend on which of the two strands is handed over as gr_seq: the same duplex, written from the other strand, can then come back with a different answer. The convention here is invariant under that swap by construction. See the 1.1.2 entry in NEWS for a worked example.

Parameter provenance and limitations

Table identifiers are retained for compatibility and are not always literal author-year citations. The installed files extdata/tm_nn_reference_audit.md and extdata/tm_nn_parameter_sources.tsv record the source and verification status of every built-in parameter row.

The audit identified unresolved numerical/model issues: the patent's GG/CG terminal-mismatch enthalpy is inconsistent with its printed free energy; RNA_NN_Chen_2012 combines refitted Watson-Crick terms with GU terms fitted against Xia (1998); and Banerjee (2020) initiation conditions in Table 2 footnotes are not reproduced by the current per-end application. The audit corrects citations without changing these numerical results. Consult the audit before using these affected models.

Author(s)

Junhui Li

References

Breslauer KJ, Frank R, Blocker H, Marky LA (1986). Predicting DNA duplex stability from the base sequence. PNAS 83:3746-3750. <doi:10.1073/pnas.83.11.3746>

Sugimoto N, Nakano S, Yoneyama M, Honda K (1996). Improved thermodynamic parameters and helix initiation factor to predict stability of DNA duplexes. NAR 24:4501-4505. <doi:10.1093/nar/24.22.4501>

Allawi HT, SantaLucia J Jr (1997). Thermodynamics and NMR of internal G.T mismatches in DNA. Biochemistry 36:10581-10594. Table 1: Watson-Crick parameters; Table 5: G.T mismatches. <doi:10.1021/bi962590c>

SantaLucia J Jr (1998). A unified view of polymer, dumbbell, and oligonucleotide DNA nearest-neighbor thermodynamics. PNAS 95:1460-1465. <doi:10.1073/pnas.95.4.1460>

SantaLucia J Jr, Hicks D (2004). The thermodynamics of DNA structural motifs. Annual Review of Biophysics and Biomolecular Structure 33:415-440. <doi:10.1146/annurev.biophys.32.110601.141800>

Freier SM et al. (1986). Improved free-energy parameters for predictions of RNA duplex stability. PNAS 83:9373-9377. <doi:10.1073/pnas.83.24.9373>

Xia T et al. (1998). Thermodynamic parameters for an expanded nearest-neighbor model for formation of RNA duplexes with Watson-Crick base pairs. Biochemistry 37:14719-14735. <doi:10.1021/bi9809425>

Chen JL et al. (2012). Testing the nearest neighbor model for canonical RNA base pairs: revision of GU parameters. Biochemistry 51:3508-3522. <doi:10.1021/bi3002709>

Sugimoto N et al. (1995). Thermodynamic parameters to predict stability of RNA/DNA hybrid duplexes. Biochemistry 34:11211-11216. <doi:10.1021/bi00035a029>

Allawi HT, SantaLucia J Jr (1998). Nearest-neighbor thermodynamics of internal A.C mismatches in DNA: sequence dependence and pH effects. Biochemistry 37:9435-9444. <doi:10.1021/bi9803729>

Allawi HT, SantaLucia J Jr (1998). Nearest neighbor thermodynamic parameters for internal G.A mismatches in DNA. Biochemistry 37:2170-2179. <doi:10.1021/bi9724873>

Allawi HT, SantaLucia J Jr (1998). Thermodynamics of internal C.T mismatches in DNA. NAR 26:2694-2701. <doi:10.1093/nar/26.11.2694>

Peyret N, Seneviratne PA, Allawi HT, SantaLucia J Jr (1999). Nearest-neighbor thermodynamics and NMR of DNA sequences with internal A.A, C.C, G.G, and T.T mismatches. Biochemistry 38:3468-3477. <doi:10.1021/bi9825091>

Watkins NE Jr, SantaLucia J Jr (2005). Nearest-neighbor thermodynamics of deoxyinosine pairs in DNA duplexes. NAR 33:6258-6267. <doi:10.1093/nar/gki918>

SantaLucia J Jr, Peyret N (2001). Method and system for predicting nucleic acid hybridization thermodynamics and computer-readable storage medium for use therein. Patent WO2001094611A2, published December 13, 2001, Tables 2-3. https://patents.google.com/patent/WO2001094611A2/en

Bommarito S, Peyret N, SantaLucia J Jr (2000). Thermodynamic parameters for DNA sequences with dangling ends. NAR 28:1929-1934. <doi:10.1093/nar/28.9.1929>

Turner DH, Mathews DH (2010). NNDB: the nearest neighbor parameter database for predicting stability of nucleic acid secondary structure. NAR 38:D280-D282. <doi:10.1093/nar/gkp892> The RNA dangling-end values are the Turner 2004 compilation: https://rna.urmc.rochester.edu/NNDB/rna_2004/rna_2004_dangling_ends.html

Weber G (2015). Optimization method for obtaining nearest-neighbour DNA entropies and enthalpies directly from melting temperatures. Bioinformatics 31:871-877. <doi:10.1093/bioinformatics/btu751>

Ferreira I, Jolley EA, Znosko BM, Weber G (2019). Replacing salt correction factors with optimized RNA nearest-neighbour enthalpy and entropy parameters. Chemical Physics 521:69-76. <doi:10.1016/j.chemphys.2019.01.016>

Basilio Barbosa V, de Oliveira Martins E, Weber G (2019). Nearest-neighbour parameters optimized for melting temperature prediction of DNA/RNA hybrids at high and low salt concentrations. Biophysical Chemistry 251:106189. <doi:10.1016/j.bpc.2019.106189>

Banerjee D et al. (2020). Improved nearest-neighbor parameters for the stability of RNA/DNA hybrids under a physiological condition. NAR 48:12042-12054. <doi:10.1093/nar/gkaa572>

Zuber J, Schroeder SJ, Sun H, Turner DH, Mathews DH (2022). Nearest neighbor rules for RNA helix folding thermodynamics: improved end effects. NAR 50:5251-5262. <doi:10.1093/nar/gkac261>

Ghosh S et al. (2020). Nearest-neighbor parameters for predicting DNA duplex stability in diverse molecular crowding conditions. PNAS 117:14194-14201. <doi:10.1073/pnas.1920886117>

Ghosh S et al. (2023). Nearest-neighbor parameters for the prediction of RNA duplex stability in diverse in vitro and cellular-like crowding conditions. NAR 51:4101-4111. <doi:10.1093/nar/gkad020>

Blake RD, Delcourt SG (1996). Thermodynamic effects of formamide on DNA stability. NAR 24:2095-2103. <doi:10.1093/nar/24.11.2095>

See Also

tm_calculate for a single entry point to the nearest-neighbor, GC-content and Wallace methods.

Examples

input_seq <- c("AAAATTTTTTTCCCCCCCCCCCCCCGGGGGGGGGGGGTGTGCGCTGC",
"AAAATTTTTTTCCCCCCCCCCCCCCGGGGGGGGGGGGTGTGCGCTGC")
seqs <- to_genomic_ranges(input_seq)
out <- tm_nn(seqs, Na=50)
out

# A parameter set fitted at a stated sodium concentration. Because Na
# matches the concentration the set was fitted at, salt correction is
# skipped automatically rather than applied on top of it.
out_ls <- tm_nn(seqs, nn_table = "RNA_DNA_NN_Weber_2019_LS", Na = 100)
out_ls$options[["Salt correction applied"]]

# -- A parameter table supplied by the user --------------------------------
# Any of nn_table, tmm_table, imm_table and de_table also accepts a matrix,
# which is how a set the package does not ship, most obviously one covering
# modified bases, is used without waiting for a new release.
#
# Start from a built-in set to get the required keys, then add a stack. The
# key naming follows the built-in convention: top strand, "/", bottom
# strand, so "MG/CG" would be a 5-methylcytosine followed by G, paired with
# CG. The values here are illustrative and are NOT measured parameters.
tbl <- TmCalculator:::get_table("DNA_NN_SantaLucia_2004")
tbl <- rbind(tbl, "MG/CG" = c(-9.1, -24.0))

# Optional attributes. "reference" names the built-in whose key set must be
# present, which matters for RNA and hybrid tables because the default
# reference is a DNA/DNA set. "salt_mM" marks the table as fitted at a
# stated sodium concentration, so that it suppresses the salt correction at
# that concentration exactly as the built-in condition-specific sets do.
attr(tbl, "reference") <- "DNA_NN_SantaLucia_2004"

out_user <- tm_nn(seqs, nn_table = tbl, Na = 50)
out_user$options[["Thermodynamic NN values"]]
# "user-supplied (reference: DNA_NN_SantaLucia_2004)"

# Passing a built-in table back in through this route changes nothing: the
# supplied table is reordered to the reference key order before use, so row
# order carries no information.
identical(tm_nn(seqs, nn_table = "DNA_NN_SantaLucia_2004")$gr$Tm,
          tm_nn(seqs,
                nn_table = TmCalculator:::get_table("DNA_NN_SantaLucia_2004")
                )$gr$Tm)

# A table missing a key is refused rather than tolerated. The compiled core
# resolves each stack by name, so an absent key would contribute zero
# enthalpy and entropy to every sequence containing that step, silently.
try(tm_nn(seqs, nn_table = tbl[setdiff(rownames(tbl), "AA/TT"), ]))
out_ls$options[["Parameter set fitted at [Na+] (mM)"]]

TmCalculator documentation built on Oct. 5, 2026, 5:08 p.m.