________
POPULATE USER PROJECT DIRECTORY WITH INPUT FILES

The following PDB format files were created in directory "/up2/exp".

Created by editing PDB entry 2QKB.
zcd.pdb       (human RNase H1 catalytic domain)
zcd_dr.pdb    (human RNase H1 catalytic domain +19-mer RNA/DNA hybrid)

The mutation D210N in the experimental structure is reversed.  For
nucleic acids, the grouping of atoms into residues, residue names,
residue order, atom names, and atom order are changed from PDB format
to the alternative format used by the "ereg" program.  More
specifically, each nucleotide residue in the PDB format is replaced by
an ordered triple of residues (pentose ring, nucleic acid base,
phosphate linkage) in the "ereg" format.  A summary of the "ereg"
format for input of nucleic acids is contained in file
"/up2/exp/zzatm".

Created by editing PDB entry 3BSU.
zhbd.pdb      (human RNase H1 hybrid binding domain)
zhbd_dr.pdb   (human RNase H1 hybrid binding domain +12-mer RNA/DNA hybrid)

Created by editing PDB entry 2KYD.
temrna12a.pdb  (12-mer double stranded RNA)

Created by editing PDB entry 355D.
temdna12b.pdb  (12-mer double stranded DNA)

Created by editing PDB entry 2MKN.
temrna19a.pdb  (19-mer double stranded RNA)

Created by editing PDB entry 5BT2.
temdna19b.pdb  (19-mer double stranded DNA)

The following protein models and protein +oligonucleotide models were
created in directory "/up2/exp".

RNaseH1_cd.pdb      (zcd)
RNaseH1_cd_dd.pdb   (zcd_dr with residue edits R-->D U-->T)
RNaseH1_cd_dr.pdb   (zcd_dr)
RNaseH1_cd_rr.pdb   (zcd_dr with residue edits D-->R T-->U)
RNaseH1_hbd.pdb     (zhbd)
RNaseH1_hbd_dd.pdb  (zhbd_dr with residue edits R-->D U-->T)
RNaseH1_hbd_dr.pdb  (zhbd_dr)
RNaseH1_hbd_rr.pdb  (zhbd_dr with residue edits D-->R T-->U)

In models for which the parent hybrid RNA/DNA double stranded
oligonucleotide has been edited to DNA/DNA or RNA/RNA, the
conformation remains the hybrid form, referred to in this project as
the H-form, intermediate between the A-form preferred by RNA and the
B-form preferred by DNA.

________
GENERATE UNBOUND NUCLEIC ACIDS IN HELIX FORMS A, B, AND H

Although the sequences of nucleic acid bases in the unbound
oligonucleotides temrna12a, temrna12b, temrna19a, and temrna19b differ
from the sequences in the RNA/DNA hybrids bound to RHase H1 domains,
these unbound structures will be used as templates for building models
for other sequences in A-form and B-form conformations.

Local energy minimization of the unbound oligonucleotide structures
improves the base pairing geometry.

Summary of the "ereg" command.
|SYNTAX
|  ereg up2 MOL
|INPUT FILES                   OUTPUT FILES
|/up2/exp/MOL.pdb              /up2/dgn/ereg.MOL
|                              /up2/seq/seq.MOL
|                              /up2/tor/tor.MOL.CNF
|                              /up2/car/MOL.CNF.pdb

From directory "/src/str", Execute commands:
% ereg up2 temrna12a
% ereg up2 temdna12b
% ereg up2 temrna19a
% ereg up2 temdna19b
% cp ../../up2/car/temrna12a.02.pdb ../../up2/exp/tem12aform.pdb
% cp ../../up2/car/temdna12b.06.pdb ../../up2/exp/tem12bform.pdb
% cp ../../up2/car/temrna19a.04.pdb ../../up2/exp/tem19aform.pdb
% cp ../../up2/car/temdna19b.05.pdb ../../up2/exp/tem19bform.pdb

The following unbound oligonucleotide models were created using
homology modeling in directory "/up2/exp".

12 nucleotide double stranded oligos
dd12a.pdb   (A-form DNA/DNA homology model from tem12aform)
dd12b.pdb   (B-form DNA/DNA homology model from tem12bform)

dr12a.pdb   (A-form RNA/DNA homology model from tem12aform)
dr12b.pdb   (B-form RNA/DNA homology model from tem12bform)

rr12a.pdb   (A-form RNA/RNA homology model from tem12aform)
rr12b.pdb   (B-form RNA/RNA homology model from tem12bform)

19 nucleotide double stranded oligos
dd19a.pdb   (A-form DNA/DNA homology model from tem19aform)
dd19b.pdb   (B-form DNA/DNA homology model from tem19bform)

dr19a.pdb   (A-form RNA/DNA homology model from tem19aform)
dr19b.pdb   (B-form RNA/DNA homology model from tem19bform)

rr19a.pdb   (A-form RNA/RNA homology model from tem19aform)
rr19b.pdb   (B-form RNA/RNA homology model from tem19bform)

Summary of the "greg" and "prof" commands.
|SYNTAX
|  greg up2 TEM
|INPUT FILES                   OUTPUT FILES
|/up2/exp/TEM.pdb              /up2/dgn/greg.TEM
|                              /up2/seq/seq.TEM
|                              /up2/tor/tor.TEM.00
|                              /up2/car/TEM.00.pdb
|SYNTAX
|  prof up2 TEM 00
|INPUT FILES                   OUTPUT FILES
|/up2/seq/seq.TEM              /up2/dgn/prof.TEM.00
|/up2/tor/tor.TEM.00           /up2/car/dprof.TEM.00
|/up2/car/TEM.00.pdb

Execute commands:
% greg up2 tem12aform
% greg up2 tem12bform
% prof up2 tem12aform 00
% prof up2 tem12bform 00

Create files:
% up2/seq/seq.dd12b
% up2/seq/seq.dr12b
% up2/seq/seq.rr12b
% up2/seq/seq.rr12a
% up2/seq/seq.dr12a
% up2/seq/seq.dd12a
% up2/arg/exptstructs.12bform
% up2/arg/exptstructs.12aform

Summary of the "hlog" command.
|SYNTAX
|  hlog up2 MOL GRP
|INPUT FILES                   OUTPUT FILES
|/up2/arg/exptstructs.GRP      /up2/dgn/hlog.MOL.GRP
|/up2/seq/seq.MOL              /up2/car/TEM.GRP.pdb
|/up2/seq/seq.TEM              /up2/tor/tor.MOL.GRP
|/up2/tor/tor.TEM.00           /up2/car/MOL.GRP.pdb
|/up2/car/TEM.00.pdb           /up2/car/dprof.MOL.GRP
|/up2/car/dprof.TEM.00         /up2/stp/stp.hot.MOL

Execute commands:
% hlog up2 dd12b 12bform
% hlog up2 rr12a 12aform
% hlog up2 dr12b 12bform
% hlog up2 rr12b 12bform
% hlog up2 dr12a 12aform
% hlog up2 dd12a 12aform
% mv ../../up2/car/dd12a.12aform.pdb ../../up2/exp/dd12a.pdb
% mv ../../up2/car/dd12b.12bform.pdb ../../up2/exp/dd12b.pdb
% mv ../../up2/car/dr12a.12aform.pdb ../../up2/exp/dr12a.pdb
% mv ../../up2/car/dr12b.12bform.pdb ../../up2/exp/dr12b.pdb
% mv ../../up2/car/rr12a.12aform.pdb ../../up2/exp/rr12a.pdb
% mv ../../up2/car/rr12b.12bform.pdb ../../up2/exp/rr12b.pdb

Execute commands:
% greg up2 tem19aform
% greg up2 tem19bform
% prof up2 tem19aform 00
% prof up2 tem19bform 00

Create files:
% up2/seq/seq.dd19b
% up2/seq/seq.dr19b
% up2/seq/seq.rr19b
% up2/seq/seq.rr19a
% up2/seq/seq.dr19a
% up2/seq/seq.dd19a
% up2/arg/exptstructs.19bform
% up2/arg/exptstructs.19aform

Execute commands:
% hlog up2 dd19b 19bform
% hlog up2 rr19a 19aform
% hlog up2 dr19b 19bform
% hlog up2 rr19b 19bform
% hlog up2 dr19a 19aform
% hlog up2 dd19a 19aform
% mv ../../up2/car/dd19a.19aform.pdb ../../up2/exp/dd19a.pdb
% mv ../../up2/car/dd19b.19bform.pdb ../../up2/exp/dd19b.pdb
% mv ../../up2/car/dr19a.19aform.pdb ../../up2/exp/dr19a.pdb
% mv ../../up2/car/dr19b.19bform.pdb ../../up2/exp/dr19b.pdb
% mv ../../up2/car/rr19a.19aform.pdb ../../up2/exp/rr19a.pdb
% mv ../../up2/car/rr19b.19bform.pdb ../../up2/exp/rr19b.pdb

The following unbound oligonucleotide models were created in directory
"/up2/exp" by removing the protein domains from the above protein
+oligonucleotide models.

12 nucleotide double stranded oligos:
dd12h.pdb   (H-form DNA/DNA isolated from RNaseH1_hbd_dd)
dr12h.pdb   (H-form RNA/DNA isolated from RNaseH1_hbd_dr)
rr12h.pdb   (H-form RNA/RNA isolated from RNaseH1_hbd_rr)

19 nucleotide double stranded oligos:
dd19h.pdb   (H-form DNA/DNA isolated from RNaseH1_cd_dd)
dr19h.pdb   (H-form RNA/DNA isolated from RNaseH1_cd_dr)
rr19h.pdb   (H-form RNA/RNA isolated from RNaseH1_cd_rr)

% cp RNaseH1_hbd_dd.pdb dd12h.pdb
% cp RNaseH1_hbd_dr.pdb dr12h.pdb
% cp RNaseH1_hbd_rr.pdb rr12h.pdb
% cp RNaseH1_cd_dd.pdb dd19h.pdb
% cp RNaseH1_cd_dr.pdb dr19h.pdb
% cp RNaseH1_cd_rr.pdb rr19h.pdb

These files are edited to include only the oligonucleotide.

________
MINIMIZE ENERGY OF UNBOUND 12-NUCLEOTIDE DOUBLE STRANDS IN A, B, AND H
FORMS

PDB entry 3BSU provides a crystal structure for a 12 nucleotide double
stranded RNA/DNA oligonucleotide bound to the hybrid binding domain of
RNase H1.  Using the sequence of nucleic acid bases from the crystal
structure, we model 3 compositions for the oligonucleotide DNA/DNA,
RNA/DNA, and RNA/RNA in both the bound and unbound states.  In the
unbound state, for each composition, we model 3 conformations for the
double helical structure: B-form, A-form, and H-form.  For each of the
3 double strand compositions, local energy minimizations enable a
comparison of predicted stability over the 3 helix forms.

Execute commands:
% ereg up2 dd12b
% ereg up2 dd12h
% ereg up2 dd12a
% ereg up2 dr12b
% ereg up2 dr12h
% ereg up2 dr12a
% ereg up2 rr12b
% ereg up2 rr12h
% ereg up2 rr12a

Table 1: Energy (kcal/mol) and RMSD (Angstrom) for 12-nucleotide
double stranded DNA/DNA, RNA/DNA, and RNA/RNA following local energy
minimizations starting from helical forms B, H, and A.
____________________________________________________________
  Energy[ RMSD]
_______________
composition               helix form
                 B             H             A
________  _____________ _____________ _____________
 DNA/DNA   396.5[ 1.06]  386.5[ 1.36]  418.9[  .76]

 RNA/DNA   522.4[  .90]  448.7[ 2.44]  480.0[  .50]

 RNA/RNA   642.2[ 2.23]  534.6[ 1.49]  532.8[ 1.16]
____________________________________________________________

For the RNA/DNA composition, model energies predict the most stable
conformation is the H-form helix.  Although the DNA/DNA composition
alters the relative energies between the A and B forms toward the
B-form, the H-form remains favored.  For the RNA/RNA composition,
model energy strongly favors the A-form helix over the B-form.

Because the starting structures for the H-form conformations are taken
from a crystal structure rather than generated by homology modeling,
possibly favoring the H-form conformation, we searched, using the
"ptra" command, for alternative local energy minima in the regions of
the B, H, and A form structures.

From directory "/up2/car", execute commands:
% cp dd12b.06.pdb dd12b.s000.pdb
% cp dd12h.06.pdb dd12h.s000.pdb
% cp dd12a.03.pdb dd12a.s000.pdb
% cp dr12b.05.pdb dr12b.s000.pdb
% cp dr12h.08.pdb dr12h.s000.pdb
% cp dr12a.03.pdb dr12a.s000.pdb
% cp rr12b.09.pdb rr12b.s000.pdb
% cp rr12h.06.pdb rr12h.s000.pdb
% cp rr12a.06.pdb rr12a.s000.pdb

From directory "/up2/tor", execute commands:
% cp tor.dd12b.06 tor.dd12b.s000
% cp tor.dd12h.06 tor.dd12h.s000
% cp tor.dd12a.03 tor.dd12a.s000
% cp tor.dr12b.05 tor.dr12b.s000
% cp tor.dr12h.08 tor.dr12h.s000
% cp tor.dr12a.03 tor.dr12a.s000
% cp tor.rr12b.09 tor.rr12b.s000
% cp tor.rr12h.06 tor.rr12h.s000
% cp tor.rr12a.06 tor.rr12a.s000

From directory "/src/str", execute commands:
% ptra up2 dd12b s000 full
% ptra up2 dd12h s000 full
% ptra up2 dd12a s000 full
% ptra up2 dr12b s000 full
% ptra up2 dr12h s000 full
% ptra up2 dr12a s000 full
% ptra up2 rr12b s000 full
% ptra up2 rr12h s000 full
% ptra up2 rr12a s000 full

Summary of the "ptra" command.
|SYNTAX
|  ptra up2 MOL CNF SUB
|INPUT FILES                   OUTPUT FILES
|/up2/seq/seq.MOL              /up2/dgn/ptra.MOL.CNF.SUB
|/up2/tor/tor.MOL.CNF          /up2/car/MOL.t???.pdb
|/up2/stp/stp.SUB.MOL          /up2/tor/tor.MOL.t???

Table 2: Energy (kcal/mol) for 12-nucleotide double stranded DNA/DNA,
RNA/DNA, and RNA/RNA following trajectory searches starting from the
endpoints of local energy minimization for helical forms B, H, and A.
____________________________________________________________
  Energy
_______________
composition               helix form
                 B             H             A
________  _____________ _____________ _____________
 DNA/DNA      377.3         371.3         382.0

 RNA/DNA      512.3         448.0         471.2

 RNA/RNA      641.1         531.0         530.2
____________________________________________________________

By lowering energies of the B and A-forms more than the H-form, these
trajectory searches improve consistency between predicted and observed
structure preferences for the DNA/DNA and RNA/RNA compositions.  The
Watson-Crick base pairing of the double helices is, in all cases,
preserved in the lowest energy conformations generated by the
trajectory.

________
MINIMIZE ENERGY OF UNBOUND 19-NUCLEOTIDE DOUBLE STRANDS IN A, B, AND H
FORMS

PDB entry 2QKB provides a crystal structure for a 19 nucleotide double
stranded RNA/DNA oligonucleotide bound to the catalytic domain of
RNase H1.  Using the sequence of nucleic acid bases from the crystal
structure, we model 3 compositions for the oligonucleotide DNA/DNA,
RNA/DNA, and RNA/RNA in both the bound and unbound states.  In the
unbound state, for each composition, we model 3 conformations for the
double helical structure: B-form, A-form, and H-form.  For each of the
3 double strand compositions, local energy minimizations enable a
comparison of predicted stability over the 3 helix forms.

Execute commands:
% ereg up2 dd19b
% ereg up2 dd19h
% ereg up2 dd19a
% ereg up2 dr19b
% ereg up2 dr19h
% ereg up2 dr19a
% ereg up2 rr19b
% ereg up2 rr19h
% ereg up2 rr19a

Table 3: Energy (kcal/mol) and RMSD (Angstrom) for 19-nucleotide
double stranded DNA/DNA, RNA/DNA, and RNA/RNA following local energy
minimizations starting from helical forms B, H, and A.
____________________________________________________________
  Energy[ RMSD]
_______________
composition               helix form
                 B             H             A
________  _____________ _____________ _____________
 DNA/DNA   546.2[ 1.96]  539.5[ 2.69]  564.9[ 3.12]

 RNA/DNA   778.2[ 2.36]  626.9[ 1.91]  644.8[ 3.67]

 RNA/RNA   990.7[ 2.96]  814.4[ 1.79]  736.1[  .38]
____________________________________________________________

Consistent with observed structure preferences, the RNA/RNA and
RNA/DNA compositions favor the A-form and H-form structures,
respectively.  Inconsistent with observed structure preferences, the
DNA/DNA composition favors the H-form structure.

________
CALCULATE ENERGIES FOR BINDING OF DNA/DNA, RNA/DNA, AND RNA/RNA TO THE
HYBRID BINDING DOMAIN OF RNASE H1

Execute commands:
% ereg up2 RNaseH1_hbd
% ereg up2 RNaseH1_hbd_dd
% ereg up2 RNaseH1_hbd_dr
% ereg up2 RNaseH1_hbd_rr

Structure energies (kcal/mol):
  F(RNaseH1_hbd)         = -1126.4
  F(RNaseH1_hbd +DNA/DNA)=  -511.9
  F(RNaseH1_hbd +RNA/DNA)=  -445.9
  F(RNaseH1_hbd +RNA/RNA)=  -331.9

Binding energies (kcal/mol):
  F(RNaseH1_hbd +DNA/DNA) -F(DNA/DNA) -F(RNaseH1_hbd)=
    -511.9 -371.3 +1126.4=  243.2
  F(RNaseH1_hbd +RNA/DNA) -F(RNA/DNA) -F(RNaseH1_hbd)=
    -445.9 -448.0 +1126.4=  232.5
  F(RNaseH1_hbd +RNA/RNA) -F(RNA/RNA) -F(RNaseH1_hbd)=
    -331.9 -530.2 +1126.4=  264.3

The model predicts, correctly, that the RNA/DNA hybrid duplex binds
more tightly than the DNA/DNA and RNA/RNA duplexes to the hybrid
binding domain.

The positive values for calculated binding energies is a side effect
of a slight difference in the form of the Fm component between the
current protein energy function and the current nucleic acid energy
function.  The energy of the unbound hybrid binding domain
F(RNaseH1_hbd) is calculated using Fm of the protein energy function.
The energies of the hybrid binding domain bound to the double stranded
oligonucleotides, and the energies of the unbound double stranded
oligonucleotides, are calculated using Fm of the nucleic acid energy
function.

A possible source of error in these estimations of binding energy is
nonoptimal representative structures for the bound states.  We note
that, for the bound structures for which energies were evaluated, we
have not yet attempted searches for more stable conformations in the
regions of these structures.

________
CALCULATE ENERGIES FOR BINDING OF DNA/DNA, RNA/DNA, AND RNA/RNA TO THE
CATALYTIC DOMAIN OF RNASE H1

Execute commands:
% ereg up2 RNaseH1_cd
% ereg up2 RNaseH1_cd_dd
% ereg up2 RNaseH1_cd_dr
% ereg up2 RNaseH1_cd_rr

Structure energies (kcal/mol):
  F(RNaseH1_cd         )=  -3941.6
  F(RNaseH1_cd +DNA/DNA)=  -2508.7
  F(RNaseH1_cd +RNA/DNA)=  -2376.2
  F(RNaseH1_cd +RNA/RNA)=  -2190.6

Binding energies (kcal/mol):
  F(RNaseH1_cd +DNA/DNA) -F(DNA/DNA) -F(RNaseH1_cd)=
    -2508.7 -539.5 +3941.6= 893.4
  F(RNaseH1_cd +RNA/DNA) -F(RNA/DNA) -F(RNaseH1_cd)=
    -2376.2 -626.9 +3941.6= 938.5
  F(RNaseH1_cd +RNA/RNA) -F(RNA/RNA) -F(RNaseH1_cd)=
    -2190.6 -736.1 +3941.6= 1014.9

The model predicts, incorrectly, that the DNA/DNA duplex binds more
tightly than the RNA/DNA duplex to the catalytic domain.

We note that the calculated binding energies are limited by not having
used trajectory search to explore for lower energy local minima for
all of the states, both unbound and bound, used in the calculation.

These results, for both the hybrid binding and catalystic domains,
indicate some inaccuracy in this first generation energy function for
nucleic acids.  Obtaining better agreement with experimental
observations for this model system will be useful as a selective
pressure for guiding the evolution the nucleic acid energy function.
