
About the Poland service:


The Poland server will calculate thermal denaturation profiles and temperaturedependent UV absorbance or gel mobility of double stranded RNA, DNA, or RNA/DNAhybrids based on sequence input and parameter settings in the Poland request form.  Details of the Poland program are given below.


The program used in these calculations was developed by Gerhard Steger from our department, for comparing theroretical predictions to experimental data, mainly optical denaturation profiles, taken at 260 and 280 nm, and TGGE (temperature gradient gel electrophoresis) experiments.


The original version was written in VAX Fortran (VMS), using the Graphics Kernel System GKS for data presentation (nice interactive program that runs also with OpenVMS; available on request). A strict F77 textonly version was produced from the full version, for use on 'the PC' (i8088, no graphics; the minimal PC ...), but also useful on other small systems.


The HUSAR staff at the DKFZ (German Cancer Research Center in Heidelberg) did a first port to the GCG environment on their Convex system, which they passed back to us for further improvement. This GCG version, now running on a DEC AXP system (OpenVMS) and GCG V7.1, is available on request.


The WWW server version is based on the F77 version, compiled by GNU f77 on a Linux station; user interface and graphics output is produced by Tcl/Tk. The basic PostScript output is converted to GIF by ImageMagick.




Programspecific informations:


Calculation is based on D. Poland's algorithm including the modification by Fixman & Freire in the implementation described by Gerhard Steger. The Poland algorithm calculates the denaturation profile for doublestranded nucleic acid using nearestneighbor stacking interactions and loop entropy functions described in the literature.


The input data required for calculation are:
 the sequence, of course (and no default here!),
 optional mismatched positions,
 the strand concentration, affecting the dissociation temperature,
 the method to calculate the final dissociation into single strands,
 the thermodynamic parameter set (DNA/DNA low salt/RNA), and
 the temperature range in which the calculation is performed.
In case you need access to the full range of input options, more options are available to the experts.


Data sets predicted and figures drawn by the program are described below; see also for OUTPUT.
 A perspective view on the temperaturedependent denaturation profile (denaturation probability vs. sequence position vs. temperature. This plot does not include the dissociation of dsNA into singlestrands; thus it shows most clearly the relative stability of the different parts of the NA.
 The temperaturedependent relative UV hypochromicities as measured in optical melting, at wavelengths of 260 and 280 nm (282 nm in case of DNA), respectively (full hypochromicity corresponds to approx. 30% of the OD at low temperature).
 The derivative form of above hypochromicities, showing the melting temperature(s) and corresponding half witdh(s) of the transitions(s), giving hints about the transition cooperativity.
 Predicted relative gel mobility, calculated according to Lerman et al., vs. temperature for different values of the 'retardation length' parameter L_{r}. This plot can be used for direct comparison with TGGE experiments; superpositions of plots generated with or without mismatched positions given are useful as a hint whether specific mismatched duplexes could be detected among homoduplexed DNA in a mixture of sample and reference double strands having undergone a denaturationrenaturation cycle, using either perpedicular or parallel TGGE.
 A 'halfdenaturation temperature' plot showing the temperature at which each base pair has a probability of 50% to be in the open state. Similar to the threedimensional plot, this plot can be used to estimate the destabilizing effect of mismatches on the surrounding part of the sequence: a temperatureshift of the TGGE transition can be expected if the lowest melting part of the sequence is directly affected by the mismatch.


Calculations can be done for oligonucleotides (>15 bases) or long double strands (>50 bases), respectively. In the case of oligonucleotide mode, a lengthdependent correction for the strand dissociation process is applied. We do not have sufficient experimental results to stringently check for this mode to give valid results, but for the length range of about 20 nucleotides there is at least experimental evidence. Using 'oligo' mode with far longer sequences gives misleading results!




References for Poland Service:


Description of implemented programs (You should cite this reference in your publication!):
Steger, G. (1994). Nucleic Acids Res. 22, 27602768.
Thermal denaturation of doublestranded nucleic acids: prediction of temperatures critical for gradient gel electrophoresis and polymerase chain reaction.
Abstract:
A program is described which calculates the thermal stability and the denaturation behaviour of doublestranded DNAs and RNAs up to a length of 1000 base pairs. The algorithm is based on recursive generation of conditional and a priori probabilities for base stacking. Output of the program may be compared directly to experimental results; thus the program may be used to optimize the nucleic acid fragments, the primers and the experimental conditions prior to experiments like polymerase chain reactions, temperaturegradient gel electrophoresis, denaturinggradient gel electrophoresis and hybridizations. The program is available in three versions; the first version runs interactively on VAXstations producing graphics output directly, the second is implemented as part of the HUSAR package at GENIUSnet, the third runs on any computer producing text output which serves as input to available graphics programs.




Original version of algorithm:
Poland, D. (1974). Biopolymers 13, 18591871.
Recursion relation generation of probability profiles for specificsequence macromolecules with longrange correlations.
Abstract:
The problem of calculating detailed probability profiles giving the probability of each unit in the chain to be in the ordered state (and all other average quantities as well including the fraction of strand association) for specificsequence macromolecules requiring statistical weights that correlate up to the total number of units in the chain (e.g., DNA, collagen) is formulated in terms of recursion relations for appropriate a priori and conditional probabilities, thus generalizing the approach of Lacombe and Simha for nearestneighbor correlations in specific sequence macromolecules. The technique allows the probability profiles for chains of thousands of units to be calculated in minutes making no approximations in the basic model.




Fixman & Freire (1977). Biopolymers 16, 26932704.
Theory of DNA melting curves




HELP for Poland Service:


The program POLAND simulates transition curves of doublestranded nucleic acids (DNA and RNA as well as DNA/RNA hybrids).
Additional information available:
OUTPUT RELATED_PROGRAMS RESTRICTIONS ALGORITHM SUGGESTIONS PARAMETERS




The program writes it ouput in numeric format, which is converted to graphics by Tk/Tcl.
Additional information available:
General description resolution 3DPlot GelPlot MeltPlot Temp50%Plot




The primary graphics output is produced as PostScript® (vector format). That format is converted to GIF® (raster format); this is a format directly displayed by your WWW browser. The resolution of the GIF images is selectable: 72 dots per inch (72 dpi) is the standard screen resolution; 150 dpi or 300 dpi are nice for printing. But be aware on NanoWeak® systems: the higher resolutions need a lot of memory and tend to crash your system.




Probability of an open basepair is plotted as a function of position in sequence and temperature.




Relative mobility is plotted as a function of temperature for the three different stiffness parameters.




Relative hypochromicity and its derivative is plotted as a function of temperature at 260 nm and 282 nm (RNA 280 nm).
References:
for RNA:
Coutts, S.M. (1971). Biochim. Biophys. Acta 232, 94106. Thermodynamics and kinetics of GC base pairing in the isolated extra arm of serinespecific tRNA from yeast
for DNA:
Blake, R.D. & Haydock, P.V. (1979). Biopolymers 18, 30893109. Effect of sodium ion on the highresolution melting of lambda DNA




Temperature is plotted at which the corresponding base stack has a probability of 50% to be in the open state. The two horizontal lines in the plot mark the temperatur range of calculation; i.e., a curve coinciding with such a line is not valid.




The Poland program calculates the denaturation behavior of doublestranded NA.
LinAll, RNAfold, and mFold calculate the secondary structure of singlestranded (R)NA; in addition LinAll and RNAfold allow the prediction of denaturation behavior of ssRNA.




The sequence has to be shorter than 1001 nucleotides but longer than 5 nucleotides.
Valid nucleotides are A, G, C, U, and T.
Calculation of asymmetric or bulge loops is not possible.




Calculation is based on Poland's algorithm including the modifications proposed by Fixman & Freire.
With the original algorithm of Poland computing time is proportional to the sqare of the sequence length.
With the modification according to Fixman & Freire computing time is proportional to 10 times the sequence length, but it works only with loop parameters according to Poland.




Hints for combination of parameters and their values.
Additional information available:
RNA DNA RNA/DNA Ionic_strength_dependence




Thermodynamic values according to Turner et al. are ideally suited for calculation in 1 M ionic strength after correction of all DeltaS values by 1.021 and all DeltaS_{GC} values by 0.961. These corrections are equivalent to a shift in T_{m} values of A:U stacks by 7 K or 2%, respectively and of G:C/G:C stacks by +7 K or +2%, respectively.
Optimal (?) parameter combination for Turner et al.:
d 1.021 1.000 0.961 (DeltaS, DeltaS(A:U), and (DeltaS(G:C) factor)
n 1.e3 (Dissociation constant ß)
c 1.e6 (ß*c_{0} = 1E8 to 1E10)
s 1.000 (loop parameter Sigma)
l g (internal loops according to Gralla & Crothers)
t 90. 120. 0.5 (Temperature range and steps)
Optimal (?) parameter combination for Pörschke et al.:
d 1.000 1.040 0.970 (DeltaS, DeltaS(A:U), and (DeltaS(G:C) factor)
n 1.e3 (Dissociation constant ß)
c 1.e6 (ß*c_{0} = 1E9 to 1E11)
s 1.e6 (loop parameter Sigma)
l p (internal loops according to Poland)
a f (algorithm according to Fixman & Freire)
t 90. 120. 0.5 (Temperature range and steps)




Thermodynamic values according to Gotoh et al. and Klump, both, are ideally suited for calculations. The parameter set of Breslauer et al. does not fit our experiments (?). The parameter set of Allawi & SantaLucia is based on a reevaluation of all known parameter sets for DNA; i.e., this set may the optimal one. Recently, a complete parameter set was published by Blake & Delcourt; in our hands this set gives nearly identical results to that of Allawi & SantaLucia. For references to the original thermodynamic parameter sets see here.
Additional information available:
Gotoh Breslauer et al. Klump SantaLucia et al. Allawi & SantaLucia Blake & Delcourt




Optimal (?) parameter combination for Gotoh:
d 1.000 (DeltaS factor)
n 1.e3 (Dissociation constant ß)
c 1.e6 (ß*c_{0} = 1E9 to 1E11)
s 1.e3 (loop parameter Sigma)
l p (internal loops according to Poland)
a f (algorithm according to Fixman & Freire)
t 60. 80. 0.5 (Temperature range and steps)




No optimal parameter combination for Breslauer et al.




Optimal (?) parameter combination for Klump:
d 1.000 (DeltaS factor)
n 1.e3 (Dissociation constant ß)
c 1.e6 (ß*c_{0} = 1E9 to 1E11)
s 1.e3 (loop parameter Sigma)
l p (internal loops according to Poland)
a f (algorithm according to Fixman & Freire)
t 70. 90. 0.5 (Temperature range and steps)




This is the "unified parameter set"!


This set gives highly similar results to Allawi & SantaLucia




Thermodynamic values according to Sugimoto et al. (1995) in 1 M NaCl. The top strand is RNA, the bottom strand is DNA (5'r3'/3'd5'); the input sequence is the RNA strand.




Following values may be used for correction of calculated T_{m}values:
T_{m,2}  T_{m,1}
 = f(G:C)*I(G:C) + (1f(G:C))*I(A:U)
log(c_{2}/c_{1})
with T_{m} = transition (midpoint, melting) temperature
c = ionic strength (=concentration of Na ions)
f(G:C) = G:C content
I(X:Y) = dependence of ionic strength of
base pair type X:Y
DNA
I(A:T) = 18.3 °C (Owen, Hill, & Lapage (1969).
Biopolymers 7, 503516.)
I(G:C) = 11.3 °C (FrankKamenetskii (1971).
Biopolymers 10, 26232624.)
RNA
I(A:U) = 20.0 °C (Steger, Müller & Riesner (1980).
I(G:C) = 8.4 °C Biochim. Biophys. Acta 606, 274284.)




Additional information available:
BASE_STACKING_(thermodynamic_parameters) ENTROPY_CORRECTION_of_base_stacking LOOP_PARAMETERS_(thermodynamic_parameters) TEMPERATURE_RANGE_OF_CALCULATION MISMATCHED_POSITIONS_in_original_sequence CONCENTRATION_and_DISSOCIATION_CONSTANT STIFFNESS_of_nucleic_acid




You can select between eight different thermodynamic parameter sets of base stacking (for loop parameters see below):
Additional information available:
RNA DNA RNA/DNA




 for RNA in 1 M NaCl
Freier, S.M., Kierzek, R., Jaeger, J.A., Sugimoto, N., Caruthers, M.H., Neilson, T. & Turner, D.H. (1986). Proc. Natl. Acad. Sci. USA 83, 93739377. Improved freeenergy parameters for predictions of RNA duplex stability.
 for RNA in 1 M NaCl
Pörschke, D., Uhlenbeck, O.C. & Martin, F.H. (1973). Biopolymers 12, 13131335. Thermodynamics and kinetics of the helixcoil transition of oligomers containing GC base pairs.




 for DNA in 0.019 M NaCl
Gotoh, O. (1983). Adv. Biophys. 16, 152. Prediction of melting profiles and local helix stability for sequenced DNA.
 for DNA in 1 M NaCl
Breslauer, K.J., Frank, R., Bloecker, H. & Marky, L.A. (1986). Proc. Natl. Acad. Sci. USA 83, 37463750. Predicting DNA duplex stability from the base sequence.
 for DNA in 0.1 M NaCl
Klump, H.H. (1987). Canad. J. Chem. 66, 804809. Energetics of order/order transitions in nucleic acids. Klump, H. (1990). in LandoltBörnstein, New Series, Group VII Biophysics, Vol. 1 Nucleic Acids, Subvol. c Spectroscopic and Kinetic Data, Physical Data I, (W. Saenger, ed.), SpringerVerlag Berlin, p. 244245. Calorimetric studies on DNAs and RNAs.
 for DNA in 1 M NaCl
SantaLucia, J. Jr., Allawi, H.T. & Seneviratne, P.A. (1996). Biochemistry 35, 35553562. Improved nearestneighbor parameters for predicting DNA duplex stability.
 for DNA in 1 M NaCl
Allawi, H.T. & SantaLucia, J. Jr. (1997). Biochemistry 36, 1058110594. Thermodynamics and NMR of Internal G·T Mismatches in DNA.
 for DNA in 0.075 M NaCl
Blake, R.D. & Delcourt, S.D. (1998). Nucleic Acids Res. 26, 33233332. Thermal stability of DNA. plus Erratum in Nucleic Acids Res. 27,3 plus personal communication


 for RNA/DNA hybrids in 1 M NaCl
Sugimoto, N., Nakano, S., Katoh, M., Matsumura, A., Nakamuta, H., Ohmichi, T., Yoneyama, M. & Sasaki, M. (1995). Biochemistry 34, 1121111216. Thermodynamic parameters to predict stability of RNA/DNA hybrid duplexes.




(Option not available by WWW)
DeltaS values of base stacking may be corrected by factors in order to simulate deviating ionic strengths. The Delta S values of the thermodynamic parameters are multiplied with these factors. The first is used for correction of all Delta S values, the second only for A:U stacks, the third only for G:C stacks. Different values are used as defaults in dependence on the chosen thermodynamic parameter set.




(Option not available by WWW; i.e., Sigma is fixed to 1.e3, and loop entropy is calculated according to Poland.)
Loops which appear during denaturation by internal base stack opening may be calculated by three different methods:
 l p ==> DeltaS(loop) = SIGMA*(loop+1)**1.75
(according to Poland or Fixman & Freire)
Use only with stacking parameters according to
Pörschke et al. (p p),
Gotoh (p g), or
Klump (p k).
 l g ==> DeltaS(loop) = SIGMA*DeltaS(loop)
(according to Gralla & Crothers)
Use only with stacking parameters according to
Turner et al. (p t).
 l t ==> DeltaS(loop) = SIGMA*DeltaS(loop)
(according to Turner et al.)
Use only with stacking parameters according to
Turner et al. (p t).
Therefore, Sigma influences the cooperativity and the half width of each transition.
With 'a f' you can change the default algorithm (only in case 'l p'). With the original algorithm of Poland (default), computing time is proportional to the sqare of the sequence length. With 'a f' the modified algorithm of Fixman & Freire is used which results in computing time proportional to 10 times the sequence length but works only with loop parameters according to Poland (l p) up to a sequence length of 1000 base pairs.
References:
Poland, D. (1974) Biopolymers 13, 18591871. Recursion Relation Generation of Probability Profiles for SpecificSequence Macromolecules with LongRange Correlations.
Fixman and Freire (1977) Biopolymers 16, 26932704. Theory of DNA melting curves.
Gralla, J. & Crothers, D.M. (1973) J. Mol. Biol. 78, 301319. Free energy of imperfect nucleic acid helices. III. Small internal loops resulting from mismatches.
Freier, S.M., Kierzek, R., Jaeger, J.A., Sugimoto, N., Caruthers, M.H., Neilson, T. & Turner, D.H. (1986) Proc. Natl. Acad. Sci. USA 83, 93739377. Improved freeenergy parameters for predictions of RNA duplex stability.




The temperature range for calculations has to be adapted to the other parameters; thus see the topic Suggestions. Not more than 150 and not less than 5 temperature points are allowed for a single calculation.




Mismatches are given as a commaseparated list of sequence positions; f.e. m 2,3,111 specifies mismatched 'base pairs' at positions 2, 3, and 111. If the mismatch is longer than a 'base pair', the position of each base has to be given separately. The sequence position of the mismatched base pair(s) may be given in any order. If the mismatch is either at the first or at the last position of the sequence, shorten the sequence by one base pair at the appropriate end instead of specifying a mismatch position. Calculation of asymmetric or bulge loops is not possible; these have to be modeled by larger mismatches (internal loops).




c 1.e6 Concentration of single strands C_{0}
n 1.e3 Dissociation constant ß
ß*c_{0} influences temperature T_{m} and half width of the second order transition, i.e. the strand separation. The dissociation constant ß has to be in the range 1. => ß => 1.E5., and the strand concentration has to be in the range 1. => C_{0} => 1.E13. If case of short oligonucleotides, ß might be calculated according to
Benight, A.S. & Wartell, R.M. (1983). Biopolymers 22, 14091425. Influence of basepair changes and cooperativity parameters on the melting curves of short DNAs.
and
Benight, A.S., Wartell, R.M. & Howell, D.K. (1981). Nature 289, 203205. Theory agrees with experimental thermal denaturation of short DNA restriction fragments.




Stiffness of nucleic acid or gel pore size is given f.e. with l 40 90 200.
References:
Lerman, L.S., Fischer, S.G., Hurley, I., Silverstein, K. & Lumelsky, N. (1984). Ann. Rev. Biophys. Bioeng. 13, 399423.
Fischer, S.G. & Lerman. L.S. (1982). Proc. Natl. Acad. Sci. USA 80, 15791583.
Riesner, D., Henco, K. & Steger, G. (1991). In: Advances in Electrophoresis, Vol. 4 (Chrambach, A., Dunn, M.J. & Radola, B.J., eds.) VCH Verlagsgesellschaft, Weinheim, pp. 169250. Temperaturgradient gel electrophoresis: A method for the analysis of conformational transitions and mutations in nucleic acids and proteins
