分子动力学模拟学习3-Gromacs数据处理
1. 周期性边界处理：
gmx trjconv -s md.tpr -f md.xtc -o md_noPBC.xtc -pbc mol -center #选择1("Protein")作为置中的组分，选择0("System")作为输出2. 计算RMSD：
gmx make_ndx -f complex.gro -o rmsd_index.ndx #新建用于计算RMSD的ndx文件，并在此运行下输入下列命令
4 & r1-303 #新建4 & 待分析氨基酸
name 28 Enzyme_backbone #将新建组命名为Enzyme_backbone
28 | r 270 | r275 #再新建Enzyme_backbone | 需要额外添加的氨基酸组，如非标准氨基酸（因为这样包括了非标准氨基酸的全部原子而只有非骨架原子，并不确定能否这样做，当然无需计算backbone而计算全部氨基酸的RMSD则没有这种情况）
q #保存并退出
gmx rms -n rmsd_index.ndx -s md.tpr -f md_noPBC.xtc -o rmsd.xvg -tu ns #期间选择Enzyme_backbone | r 270 | r 275组，以ns的时间间隔运行RMSD计算
3. 计算RMSF：
gmx make_ndx -f complex.gro -o rmsf_index.ndx #和上面类似，新建ndx文件
3 * r 1-303 #新建待分析氨基酸组ca组
name 28 Enzyme_ca #重命名
28 | r 270 | 275 #新建包括非标准氨基酸的组
q #保存并退出
gmx rmsf -n rmsf_index.ndx -f md_noPBC.xtc -s md.tpr -o rmsf-per-residue.xvg -ox average.pdb -oq bfactors-residue.pdb -res #选择组29，计算个氨基酸的RMSF4. 聚类分析及找代表性结构：
gmx make_ndx -f complex.gro -o cluster_index.ndx #新建进行cluster的ndx文件，并在其运行下执行一下命令
r 1-311 #新建待分析的氨基酸组
name 28 Enzyme #重命名新建组28为Enzyme
q #保存并退出
gmx cluster -f md_noPBC.xtc -s md.tpr -cutoff 0.15 -method gromos -n cluster_index.ndx -cl cluster.pdb -g cluster.log #选择组28，选择合适的cutoff（此例中时0.15）保证获得的cluster数量合适，最终获得cluster.pdb，第一个构象最多的cluster就是代表性构象5. gmx_mmpbsa计算结合能
gmx trjconv -f md_noPBC.xtc -o trj_50-100ns.xtc -b 50000 -e 100000 #轨迹采样，本例共模拟了100 ns，取50-100ns
gmx check -f trj_50-100ns.xtc #轨迹检查
gmx make_ndx -f complex.gro -o mmpbsa_index.ndx #新建用于计算的ndx组，包括Enzyme组和ligand组，具体操作和上述相似，就不多说了
之后计算主要依赖李继存老师的gmx_mmpbsa.bsh的脚本，可以参考以下内容：结合自由能计算MM/PBSA
本例中的 gmx_mmpbsa.bsh脚本内容如下,可供参考：
echo -e "\
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> gmx_mmpbsa <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Jicun Li <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
>>>>>>>>>>>>>>>>>>>>>>>>>> 2022-07-14 11:17:41 <<<<<<<<<<<<<<<<<<<<<<<<<\n
>> Usage: gmx_mmpbsa [-f *.xtc ] [-s *.tpr ] [-n *.ndx ]
[-pro PROTEIN] [-lig ]
[-cou dh ] [-ts ie ] [-cas ]\n
>> Default: gmx_mmpbsa -f traj.xtc -s topol.tpr -n index.ndx
-pro Protein -lig Ligand
-ts ie
--------------------------------------------------------------------------------
>> Option:
-f: trajectory file
-s: topology file
-n: index file
-pro: index group name of protein
-lig: index group name of ligand, can be ignored using none
-cou: dh: calculate MM(COU) with Debye-Huckel screening method
-ts: ie: calculate entropy with interaction entropy method
-cas: select res for CAS, begin:end:step or number, using global resnum
Other: change them in the script directly
--------------------------------------------------------------------------------
>> Warning:
!!! 1. make sure gmx is available and version is consistent
!!! 2. gawk version > 3; macOS awk version > 2018
!!! 3. watch _pid.pdb to make sure PBC is correct
--------------------------------------------------------------------------------
>> Log:
TODO: parallel APBS, focus
2022-07-14: fix issues with Chinease in PBSAset
2022-07-08: fix bug, No PB SA value
2022-07-01: fix bug for awk too many open files
2022-06-18: fix bug for ndx when Lig is 1st
2022-04-19: fix bug for CAS
remove -com option
2022-02-09: CAS except GLY, PRO, CYX
fix radius bug, use AMBER mBondi
change to AMBER PB4
2021-11-25: keep original PDB residue name
output avaraged contibutions
clean output files
2021-05-17: add systime and fflush, only for gawk
2021-05-16: revise cmd for all Linux
see https://github.com/Jerkwin/gmxtool/issues/2
2021-03-14: add -cou, -ts option, see 10.1088/0256-307X/38/1/018701
change default PB method to npbe
2021-03-03: fix bug: awk will exit for multiple system call
2020-11-15: fix bug for resID >=1000, for awk 3.x gsub /\s/
2020-06-02: fix bug of withLig
2020-06-01: fix bug of -skip
2020-05-27: fix bug for sdie
2020-05-26: fix bug for RES name
2020-04-03: use C6, C12 directly
2020-01-08: support protein only
2019-12-24: fix bug for small time step
2019-12-10: fix bug for OPLS force field
2019-11-17: fix bug for c6, c12 of old version tpr
2019-11-03: fix bug for time stamp
2019-09-19: push to gmxtool
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>\n"
################################################################################
# 设置运行环境, 计算参数
# setting up environmets and parameters
################################################################################
trj=trj_50-100ns.xtc # 轨迹文件 trajectory file
tpr=md.tpr # tpr文件 tpr file
ndx=mmpbsa_index.ndx # 索引文件 index file
pro=Enzyme # 蛋白索引组 index group name of protein
lig=ligand # 配体索引组 index group name of ligand
step=0 # 从第几步开始运行 step number to run
# 1. 预处理轨迹: 复合物完整化, 团簇化, 居中叠合, 然后生成pdb文件
# 2. 获取每个原子的电荷, 半径, LJ参数, 然后生成qrv文件
# 3. MM-PBSA计算: pdb->pqr, 输出apbs, 计算MM, APBS
# 1. pre-processe trajectory, whole, cluster, center, fit, then generate pdb file
# 2. abstract atomic parameters, charge, radius, C6/C12, then generate qrv file
# 3. run MM-PBSA, pdb->pqr, apbs, then calculate MM, PB, SA
gmx='gmx' # /path/to/GMX/bin/gmx_mpi
dump="$gmx dump" # gmx dump
trjconv="$gmx trjconv -dt 1000" # gmx trjconv, use -b -e -dt, NOT -skip
apbs='apbs' # APBS(Windows), USE "/", NOT "\"
pid=pid # 输出文件($$可免重复) prefix of the output files($$)
err=_$pid.err # 屏幕输出文件 file to save the message from the screen
qrv=_$pid.qrv # 电荷/半径/VDW参数文件 to save charge/radius/vdw parmeters
pdb=_$pid.pdb # 轨迹构型文件 to save trajectory
radType=1 # 原子半径类型 radius of atoms (0:ff; 1:mBondi)
radLJ0=1.2 # 用于LJ参数原子的默认半径(A, 主要为H) radius when LJ=0 (H)
meshType=0 # 网格大小 mesh (0:global 1:local)
gridType=1 # 格点大小 grid (0:GMXPBSA 1:psize)
cfac=3 # 分子尺寸到粗略格点的放大系数
# Factor to expand mol-dim to get coarse grid dim
fadd=10 # 分子尺寸到细密格点的增加值(A)
# Amount added to mol-dim to get fine grid dim (A)
df=0.5 # 细密格点间距(A) The desired fine mesh spacing (A)
# 极性计算设置(Polar)
PBEset='
temp 298.15 # 温度
pdie 2 # 溶质介电常数
sdie 78.54 # 溶剂介电常数, 真空1, 水78.54
npbe # PB方程求解方法, lpbe(线性), npbe(非线性), smbpe(大小修正)
bcfl mdh # 粗略格点PB方程的边界条件, zero, sdh/mdh(single/multiple Debye-Huckel), focus, map
srfm smol # 构建介质和离子边界的模型, mol(分子表面), smol(平滑分子表面), spl2/4(三次样条/7阶多项式)
chgm spl4 # 电荷映射到格点的方法, spl0/2/4, 三线性插值, 立方/四次B样条离散
swin 0.3 # 立方样条的窗口值, 仅用于 srfm=spl2/4
srad 1.4 # 溶剂探测半径
sdens 10 # 表面密度, 每A^2的格点数, (srad=0)或(srfm=spl2/4)时不使用
ion charge 1 conc 0.15 radius 0.95 # 阳离子的电荷, 浓度, 半径
ion charge -1 conc 0.15 radius 1.81 # 阴离子
calcforce no
calcenergy comps'
# 非极性计算设置(Apolar/Non-polar)
PBAset='
temp 298.15 # 温度
srfm sacc # 构建溶剂相关表面或体积的模型
swin 0.3 # 立方样条窗口(A), 用于定义样条表面
# SASA
srad 1.4 # 探测半径(A)
gamma 1 # 表面张力(kJ/mol-A^2)
#gamma const 0.0226778 3.84928
#gamma const 0.027 0
#gamma const 0.0301248 0 # AMBER-PB4 .0072*cal2J 表面张力, 常数
press 0 # 压力(kJ/mol-A^3)
bconc 0 # 溶剂本体密度(A^3)
sdens 10
dpos 0.2
grid 0.1 0.1 0.1
# SAV
#srad 1.29 # SAV探测半径(A)
#press 0.234304 # 压力(kJ/mol-A^3)
# WCA
#srad 1.25 # 探测半径(A)
#sdens 200 # 表面的格点密度(1/A)
#dpos 0.05 # 表面积导数的计算步长
#bconc 0.033428 # 溶剂本体密度(A^3)
#grid 0.45 0.45 0.45 # 算体积分时的格点间距(A)
calcforce no
calcenergy total'
################################################################################
# 检查 gmx, apbs 是否可以运行
# check gmx, apbs availability
################################################################################
str=$($gmx --version | grep -i "GROMACS version")
[[ $step -lt 3 && -z "$str" ]] && { echo -e "!!! ERROR !!! GROMACS NOT available !\n"; exit; }
str=$($apbs --version 2>&1 | grep -i "Poisson-Boltzmann")
[[ -z "$str" ]] && { echo -e "!!! WARNING !!! APBS NOT available !\n"; }
################################################################################
# 解析命令行参数
# parse command line options
################################################################################
useDH=1; useTS=1; isCAS=0
cas="$*";
cas=${cas##*-cas}; cas=${cas%%-*}; cas=${cas// /}
opt=($*); N=${#opt[@]}
for((i=0; i $err
ndx=_$pid.ndx
fi
echo -e ">> 0. set up environmets and parameters: OK !\n"
fi
if [[ $step -le 1 ]]; then
################################################################################
# 1. 预处理轨迹: 复合物完整化, 团簇化, 居中叠合, 然后生成pdb文件
# 请检查pdb文件确保构型PBC处理正确
# 1. pre-processe trajectory, whole, cluster, center, fit, then generate pdb file
################################################################################
trjwho=_$pid~trj_who.xtc; trjcnt=_$pid~trj_cnt.xtc; trjcls=_$pid~trj_cls.xtc
echo $com | $trjconv -s $tpr -n $ndx -f $trj -o $trjwho &>>$err -pbc whole
[[ ! -f "$trjwho" ]] && { echo -e "!!! ERROR !!! gmx trjconv Failed ! Check $err to find out why.\n"; exit; }
if [[ $withLig -eq 1 ]]; then
# usful for single protein and ligand
echo -e "$lig\n$com" | $trjconv -s $tpr -n $ndx -f $trjwho -o $trjcnt &>>$err -pbc mol -center
echo -e "$com\n$com" | $trjconv -s $tpr -n $ndx -f $trjcnt -o $trjcls &>>$err -pbc cluster
echo -e "$lig\n$com" | $trjconv -s $tpr -n $ndx -f $trjcls -o $pdb &>>$err -fit rot+trans
else
echo -e "$lig\n$com" | $trjconv -s $tpr -n $ndx -f $trjwho -o $pdb &>>$err -pbc mol -center
fi
echo -e ">> 1. preprocessing trajectory: OK !\n"
fi; if [[ $step -le 2 ]]; then
################################################################################
# 2. 获取每个原子的电荷, 半径, LJ参数, 然后生成qrv文件
# 2. abstract atomic parameters, charge, radius, C6/C12, then generate qrv file
# feel free to change radius with radType
# radType=0: radius from C6/C12, or radLJ0 if either C6 or C12 is zero
# radType=1: mBondi
################################################################################
$dump -quiet -s $tpr 2>>$err \
| awk >$qrv -v ndx=$ndx -v pro=$pro -v lig=$lig -v withLig=$withLig \
-v radType=$radType -v radLJ0=$radLJ0 '
BEGIN { RS="["
print pro, lig
while(getline < ndx) {
gsub(" ","", $1); gsub("\t","", $1)
if($1==pro) for(i=3; i<=NF; i++) ndxPro[$i+0]++
if($1==pro"]") for(i=2; i<=NF; i++) ndxPro[$i+0]++
if(withLig) {
if($1==lig) for(i=3; i<=NF; i++) ndxLig[$i+0]++
if($1==lig"]") for(i=2; i<=NF; i++) ndxLig[$i+0]++
}
}
RS="\r?\n"
nres=0
}
/#molblock/ { Ntyp=$3 }
/moltype.+=/ { Imol=$3; getline; Nmol[Imol]=$3 }
/ffparams:/ {
getline Atyp; sub(/.+=/, "", Atyp); Atyp += 0
print Atyp
getline
for(i=0; i0) {
getline
while(getline) {
if(!index($0, "ANGLES")) break
k=$NF; j=$(NF-1); i=$(NF-2)
tag=toupper(Tatm[Imol,i])
if(substr(tag,1,1)=="H") Tatm[Imol,i]=sprintf("H%s", Tatm[Imol,j])
tag=toupper(Tatm[Imol,k])
if(substr(tag,1,1)=="H") Tatm[Imol,k]=sprintf("H%s", Tatm[Imol,j])
}
}
}
END {
Ntot=0; Nidx=0
for(i=0; i0) radi=getRadi(Tatm[i, j], radType)
printf "%6d %9.5f %9.6f %6d %9.6f %9.6f %6d %s %s %-6s ", \
Nidx, Qatm[i,j], radi, Catm[i,j], Satm[i,j], Eatm[i,j], \
Ntot, Name[i]"-"n+1"."j+1, \
resName[resID[i,j]], Tatm[i, j]
if(Ntot in ndxPro) print "Pro"
if(Ntot in ndxLig) print "Lig"
}
}
}
}
}
function getRadi(tag, radType) { # mBondi from AMBER20/parmed/tools/changeradii.py
radBondi["C" ]= 1.7 ; radBondi["H" ]= 1.2
radBondi["N" ]= 1.55 ; radBondi["HC"]= 1.3
radBondi["O" ]= 1.5 ; radBondi["HN"]= 1.3
radBondi["F" ]= 1.5 ; radBondi["HP"]= 1.3
radBondi["SI"]= 2.1 ; radBondi["HO"]= 0.8
radBondi["P" ]= 1.85 ; radBondi["HS"]= 0.8
radBondi["S" ]= 1.8 # radBondi["BR"]= 1.85
radBondi["CL"]= 1.7 # radBondi["I" ]= 1.98
tag=toupper(tag)
if(length(tag)>=2) {
if(radBondi[substr(tag,1,2)]) return radBondi[substr(tag,1,2)]
else return radBondi[substr(tag,1,1)]
} else {
if(radBondi[tag]) return radBondi[tag]
else return 1.5
}
}
'
[[ ! -f "$qrv" ]] && { echo -e "!!! ERROR !!! gmx dump Failed ! Check $err to find out why.\n"; exit; }
if [[ $isCAS -eq 1 ]]; then
################################################################################
# CAS: qrv pdb
################################################################################
awk -v qrv=$qrv -v cas="$cas" -v RS="\r?\n" '
BEGIN {
txt="NME Z ALA A ARG R ASN N ASP D ASH D CYS C CYM C CYX C "
txt=txt"GLU E GLH E GLN Q GLY G HID H HIE H HIP H ILE I LEU L LYS K LYN K "
txt=txt"MET M PHE F PRO P SER S THR T TRP W TYR Y VAL V"
n=split(txt, arr)
for(i=1; i<=n/2; i++) resnm[arr[2*i-1]]=arr[2*i]
split(cas,arr,",")
for(i in arr) {
if(index(arr[i],":")) {
n=split(arr[i], arr2, ":")
s=1; if(n>2) s=arr2[3]
for(j=arr2[1]; j<=arr2[2]; j+=s) resCAS[j]=1
} else resCAS[arr[i]]=1
}
getline < qrv
getline Atyp < qrv
for(i=0; iout
getline Atyp < qrv; print Atyp>out
for(i=0; iout }
while(getline < qrv) {
res=$(NF-2); gsub(/[0-9]+/,"",res)
idx=$(NF-2); gsub(/^0*/,"",idx); gsub(/[A-Z]+/,"",idx)
atm=$(NF-1)
if(!resCAS[idx]) print >out
else {
if(res=="GLY") {
# CH3
} else if(res=="PRO") {
# HN
} else if( ( res=="SER" && atm=="OG" ) \
|| ( res=="THR" && (atm=="OG1"||atm=="CG2") ) \
|| ((res=="ILE"||res=="VAL") && (atm=="CG1"||atm=="CG2") ) \
|| ((res=="CYS"||res=="CYM"||res=="CYX") && atm=="SG" ) \
|| ( atm=="CG") ) atm="HB3"
if(atm=="N" || atm=="H" || atm=="HN" \
|| atm=="C" || atm=="O" \
|| atm=="CA" || atm=="CB" || atm=="HA" || atm=="HCA" \
|| atm=="HB" || atm=="HB1" || atm=="HB2" || atm=="HB3") {
$2=Qala[atm]; $5=Sala[atm]
$3=Rala[atm]; $6=Eala[atm]
$4=Cala[atm]; $9=sprintf("%05d%s", idx, resnm[res]"2"resnm["ALA"])
printf "%6d %9.5f %9.6f %6d %9.6f %9.6f %6d %s %s %-6s %s\n",
$1, $2, $3, $4, $5, $6, $7, $8, $9, atm, $11 >out
}
}
}
close(qrv)
}
/^ATOM/ {
ATOM=substr($0,1,6)
INDX=substr($0,7,5)+0; isCAS[INDX]=0
NAME=substr($0,13,4); name=NAME; gsub(/ /,"", name)
RES =substr($0,18,3)
CHN =substr($0,22,1); if(CHN=" ") CHN="A"
NUM =substr($0,23,4)+0
X=substr($0,31,8); X += 0
Y=substr($0,39,8); Y += 0
Z=substr($0,47,8); Z += 0
if(resCAS[NUM]) {
Rhb=1.1
if(name=="CB") { Xcb=X; Ycb=Y; Zcb=Z }
if(RES=="GLY") {
# CH3
} else if(RES=="PRO") {
# HN
} else if( ( RES=="SER" && name=="OG" ) \
|| ( RES=="THR" && (name=="OG1"||name=="CG2") ) \
|| ((RES=="ILE"||RES=="VAL") && (name=="CG1"||name=="CG2") ) \
|| ((RES=="CYS"||RES=="CYM"||RES=="CYX") && name=="SG" ) \
|| ( name=="CG") ) {
NAME=" HB "; name="HB"
dx=X-Xcb; dy=Y-Ycb; dz=Z-Zcb
r=sqrt(dx*dx+dy*dy+dz*dz)
X=Xcb+Rhb*dx/r
Y=Ycb+Rhb*dy/r
Z=Zcb+Rhb*dz/r
}
RES=resnm[RES]"2"resnm["ALA"]
if(!(name=="N" || name=="H" || name=="HN" \
|| name=="CA" || name=="HA" \
|| name=="CB" || name=="HB" || name=="HB1" || name=="HB2" || name=="HB3" \
|| name=="C" || name=="O")) next
}
printf("%-6s%5d %-4s %3s %s%4d %8.3f%8.3f%8.3f\n", \
ATOM, INDX, NAME, RES, CHN, NUM, X, Y, Z)
next
}
{print}
' $pdb > ${pdb/.pdb/_CAS.pdb}
qrv=${qrv/.qrv/_CAS.qrv}
pdb=${pdb/.pdb/_CAS.pdb}
fi
echo -e ">> 2. generate qrv file: OK !\n"
fi; if [[ $step -le 3 ]]; then
################################################################################
# 3. MM-PBSA计算: pdb->pqr, 输出apbs, 计算MM, APBS
# 3. run MM-PBSA, pdb->pqr, apbs, then calculate MM, PB, SA
################################################################################
echo -e ">> 3. run MM-PBSA calculatons"
dt=$(awk '/t=/{n++;sub(/.*t=/,"");sub(/step=.*/,"");t[n]=$0;if(n==2){print t[n]-t[1];exit}}' $pdb)
awk -v pid=_$pid -v qrv=$qrv -v dt="$dt" \
-v apbs="$apbs" -v useDH=$useDH -v useTS=$useTS \
-v PBEset="$PBEset" -v PBAset="$PBAset" \
-v meshType=$meshType -v gridType=$gridType -v gmem=$gmem \
-v fadd=$fadd -v cfac=$cfac -v df=$df \
-v withLig=$withLig -v RS="\r?\n" '
BEGIN {
getline < qrv
getline Atyp < qrv
for(i=0; i Fout"_pro.pqr"
minXpro[Nfrm]=min(minXpro[Nfrm], X-r); maxXpro[Nfrm]=max(maxXpro[Nfrm], X+r)
minYpro[Nfrm]=min(minYpro[Nfrm], Y-r); maxYpro[Nfrm]=max(maxYpro[Nfrm], Y+r)
minZpro[Nfrm]=min(minZpro[Nfrm], Z-r); maxZpro[Nfrm]=max(maxZpro[Nfrm], Z+r)
}
if(withLig) {
print txt > Fout"_com.pqr"
if(INDX in ndxLig) {
print txt > Fout"_lig.pqr"
minXlig[Nfrm]=min(minXlig[Nfrm], X-r); maxXlig[Nfrm]=max(maxXlig[Nfrm], X+r)
minYlig[Nfrm]=min(minYlig[Nfrm], Y-r); maxYlig[Nfrm]=max(maxYlig[Nfrm], Y+r)
minZlig[Nfrm]=min(minZlig[Nfrm], Z-r); maxZlig[Nfrm]=max(maxZlig[Nfrm], Z+r)
}
}
minXcom[Nfrm]=min(minXpro[Nfrm], minXlig[Nfrm]); maxXcom[Nfrm]=max(maxXpro[Nfrm], maxXlig[Nfrm])
minYcom[Nfrm]=min(minYpro[Nfrm], minYlig[Nfrm]); maxYcom[Nfrm]=max(maxYpro[Nfrm], maxYlig[Nfrm])
minZcom[Nfrm]=min(minZpro[Nfrm], minZlig[Nfrm]); maxZcom[Nfrm]=max(maxZpro[Nfrm], maxZlig[Nfrm])
minX=min(minX, minXcom[Nfrm]); maxX=max(maxX, maxXcom[Nfrm])
minY=min(minY, minYcom[Nfrm]); maxY=max(maxY, maxYcom[Nfrm])
minZ=min(minZ, minZcom[Nfrm]); maxZ=max(maxZ, maxZcom[Nfrm])
next
}
END {
close(Fname[Nfrm]"_com.pqr")
close(Fname[Nfrm]"_pro.pqr")
if(withLig) close(Fname[Nfrm]"_lig.pqr")
kJcou=1389.35457520287
Rcut=1E10 # large enough
for(i=1; i<=Npro; i++) dE[resPro[i]]=0
for(i=1; i<=Nlig; i++) dE[resLig[i]]=0
Nres=asorti(dE, Tres)
txt="#Frame "; txtDH=txt
for(i=1; i<=Nres; i++) {
ii=Tres[i]; sub(/~0+/, "~", ii)
txt = txt""sprintf("%10s", ii)
txtDH = txtDH""sprintf("%21s", ii"(with DH)")
}
print txt > pid"~resPBSA.dat"
print txt > pid"~resPBSA_PB.dat"
print txt > pid"~resPBSA_SA.dat"
if(withLig) {
print txt > pid"~resMM_VDW.dat"
print txtDH > pid"~resMM.dat"
print txtDH > pid"~resMM_COU.dat"
print txtDH > pid"~res_MMPBSA.dat"
}
print " #Frame Binding( with DH ) " \
"| MM ( with DH ) PB SA " \
"| COU ( with DH ) VDW " \
"| PBcom PBpro PBlig " \
"| SAcom SApro SAlig" >> pid"~MMPBSA.dat"
maxstr=0
for(fr=1; fr<=Nfrm; fr++) maxstr=max(maxstr, length(Fname[fr]))
if(withLig) {
for(i=1; i<=Nres; i++) {
tag=Tres[i]
dGres[tag,0]=0; dGres[tag,1]=0
dHres[tag,0]=0; dHres[tag,1]=0
MMres[tag,0]=0; MMres[tag,1]=0
COUres[tag,0]=0; COUres[tag,1]=0
VDWres[tag]=0; dPBres[tag]=0; dSAres[tag]=0
}
}
for(fr=1; fr<=Nfrm; fr++) {
t_start = systime(); fflush("")
Fout=Fname[fr]
printf " >> Frame %3d/%d: %-"maxstr"s", fr, Nfrm, Fout
txt=Fout"_pro.pqr"; if(withLig) txt=Fout"_com.pqr";
close(txt)
n=0;
while(getline < txt) { n++;
qr[n]=$2; type[n]=$3; res[n]=$4;
x[n]=$(NF-4); y[n]=$(NF-3); z[n]=$(NF-2)
resID[n]=$(NF-5); gsub(/[A-Z]+/, "", resID[n])
}
close(txt)
# MM
if(withLig) {
for(i=1; i<=Npro; i++) { dEcou[resPro[i]]=0; dEcouDH[resPro[i]]=0; dEvdw[resPro[i]]=0 }
for(i=1; i<=Nlig; i++) { dEcou[resLig[i]]=0; dEcouDH[resLig[i]]=0; dEvdw[resLig[i]]=0 }
for(i=1; i<=Npro; i++) {
ii=i+Ipro-1
qi=Qatm[qr[ii]]; ci=Catm[qr[ii]]
xi=x[ii]; yi=y[ii]; zi=z[ii]
for(j=1; j<=Nlig; j++) {
jj=j+Ilig-1; cj=Catm[qr[jj]]
r=sqrt( (xi-x[jj])^2+(yi-y[jj])^2+(zi-z[jj])^2 )
if(r Fout".apbs"
else print "read\n" \
" mol pqr "Fout"_pro.pqr\n" \
"end\n\n" > Fout".apbs"
if(meshType==0) { # GMXPBSA
if(withLig) print \
dimAPBS(Fout"_com", 1, minX, maxX, minY, maxY, minZ, maxZ), \
dimAPBS(Fout"_pro", 2, minX, maxX, minY, maxY, minZ, maxZ), \
dimAPBS(Fout"_lig", 3, minX, maxX, minY, maxY, minZ, maxZ) > Fout".apbs"
else print \
dimAPBS(Fout"_pro", 1, minX, maxX, minY, maxY, minZ, maxZ) > Fout".apbs"
} else if(meshType==1) { # g_mmpbsa
if(withLig) print \
dimAPBS(Fout"_com", 1, minXcom[fr], maxXcom[fr], minYcom[fr], maxYcom[fr], minZcom[fr], maxZcom[fr]), \
dimAPBS(Fout"_pro", 2, minXpro[fr], maxXpro[fr], minYpro[fr], maxYpro[fr], minZpro[fr], maxZpro[fr]), \
dimAPBS(Fout"_lig", 3, minXlig[fr], maxXlig[fr], minYlig[fr], maxYlig[fr], minZlig[fr], maxZlig[fr]) > Fout".apbs"
else print \
dimAPBS(Fout"_pro", 1, minXpro[fr], maxXpro[fr], minYpro[fr], maxYpro[fr], minZpro[fr], maxZpro[fr]) > Fout".apbs"
}
close(Fout".apbs")
cmd=apbs" "Fout".apbs > "Fout".out 2>&1";
cmd | getline err; close(cmd)
txt=Fout".out";
while(getline < txt ) {
if(index($0, "CALCULATION #")) {
if(index($0, "("Fout"_com")) { t=1; n=Ncom }
if(index($0, "("Fout"_pro")) { t=2; n=Npro }
if(index($0, "("Fout"_lig")) { t=3; n=Nlig }
if(index($0, "~VAC)")) t += 10
if(index($0, "~SAS)")) t += 20
while(getline < txt) {
if(t<20 && index($0, "Per-atom energies:") \
|| t>20 && index($0, "Solvent Accessible Surface Area")) break
}
for(i=1; i<=n; i++) {
getline > pid"~MMPBSA.dat"
total_time = (Nfrm-fr)*(systime() - t_start)
printf " MM-PBSA(with DH) = %9.3f(%9.3f) kJ/mol Estimated Time Remaining: ~ %0dh:%02dm:%02ds\n", \
dh[fr,0], dh[fr,1], int(total_time/3600), int(total_time%3600/60), total_time%60
fmt ="%s%9.3f %s"
fmt2="%s%9.3f(%9.3f) %s"
for(i=1; i<=Nres; i++) {
ii=""; if(i==1) ii=sprintf("%-9s", Fout)
txt=""; if(i==Nres) txt="\n"
tag=Tres[i]
if(withLig) {
printf fmt, ii, dEvdw[tag], txt > pid"~resMM_VDW.dat"
printf fmt2, ii, dEcou[tag], dEcouDH[tag], txt > pid"~resMM_COU.dat"
printf fmt2, ii, dEcou[tag]+dEvdw[tag], dEcouDH[tag]+dEvdw[tag], txt > pid"~resMM.dat"
printf fmt2, ii, dEcou[tag]+dEvdw[tag]+dPBres[tag]+dSAres[tag],
dEcouDH[tag]+dEvdw[tag]+dPBres[tag]+dSAres[tag], txt > pid"~res_MMPBSA.dat"
COUres[tag,0] += dEcou[tag]
COUres[tag,1] += dEcouDH[tag]
VDWres[tag]+= dEvdw[tag];
}
PBres[tag] += dPBres[tag]
SAres[tag] += dSAres[tag]
printf fmt, ii, preK*(dPBres[tag]+dSAres[tag]), txt > pid"~resPBSA.dat"
printf fmt, ii, preK*(dPBres[tag]), txt > pid"~resPBSA_PB.dat"
printf fmt, ii, preK*(dSAres[tag]), txt > pid"~resPBSA_SA.dat"
}
fmt="%s%6s%6s\n"
for(i=1; i<=Npro; i++) {
ii=Ipro+i-1
txt=sprintf("%-6s%5d %-4s %3s A%4d %8.3f%8.3f%8.3f", \
"ATOM", ii, type[ii], res[ii], resID[ii], x[ii], y[ii], z[ii])
tag=resPro[i]
printf fmt, txt, fixfmt(preK*dPBres[tag],6), fixfmt(preK*dPBres[tag],6) > Fout"~resPBSA_PB.pdb"
printf fmt, txt, fixfmt(preK*dSAres[tag],6), fixfmt(preK*dSAres[tag],6) > Fout"~resPBSA_SA.pdb"
if(withLig) {
printf fmt, txt, fixfmt(dEcou[tag],6), fixfmt(dEcouDH[tag],6) > Fout"~resMM_COU.pdb"
printf fmt, txt, fixfmt(dEvdw[tag],6), fixfmt(dEvdw[tag],6) > Fout"~resMM_VDW.pdb"
printf fmt, txt, fixfmt(dEcou[tag]+dEvdw[tag],6), fixfmt(dEcouDH[tag]+dEvdw[tag],6) > Fout"~resMM.pdb"
printf fmt, txt, fixfmt(dPBres[tag]+dSAres[tag],6), fixfmt(dPBres[tag]+dSAres[tag],6) > Fout"~resPBSA.pdb"
printf fmt, txt, fixfmt(dEcou[tag]+dEvdw[tag]+preK*(dPBres[tag]+dSAres[tag]),6),
fixfmt(dEcouDH[tag]+dEvdw[tag]+preK*(dPBres[tag]+dSAres[tag]),6) > Fout"~res_MMPBSA.pdb"
} else
printf fmt, txt, fixfmt(preK*(dPBres[tag]+dSAres[tag]),6),
fixfmt(preK*(dPBres[tag]+dSAres[tag]),6) > Fout"~resPBSA.pdb"
}
for(i=1; i<=Nlig; i++) {
ii=Ilig+i-1
txt=sprintf("%-6s%5d %-4s %3s A%4d %8.3f%8.3f%8.3f", \
"ATOM", ii, type[ii], res[ii], resID[ii], x[ii], y[ii], z[ii])
tag=resLig[i]
printf fmt, txt, fixfmt(dEcou[tag]+dEvdw[tag]+preK*(dPBres[tag]+dSAres[tag]),6),
fixfmt(dEcouDH[tag]+dEvdw[tag]+preK*(dPBres[tag]+dSAres[tag]),6) > Fout"~res_MMPBSA.pdb"
printf fmt, txt, fixfmt(preK*dPBres[tag],6), fixfmt(preK*dPBres[tag],6) > Fout"~resPBSA_PB.pdb"
printf fmt, txt, fixfmt(preK*dSAres[tag],6), fixfmt(preK*dSAres[tag],6) > Fout"~resPBSA_SA.pdb"
printf fmt, txt, fixfmt(dEcou[tag],6), fixfmt(dEcouDH[tag],6) > Fout"~resMM_COU.pdb"
printf fmt, txt, fixfmt(dEvdw[tag],6), fixfmt(dEvdw[tag],6) > Fout"~resMM_VDW.pdb"
printf fmt, txt, fixfmt(dEcou[tag]+dEvdw[tag],6), fixfmt(dEcouDH[tag]+dEvdw[tag],6) > Fout"~resMM.pdb"
printf fmt, txt, fixfmt(dPBres[tag]+dSAres[tag],6), fixfmt(dPBres[tag]+dSAres[tag],6) > Fout"~resPBSA.pdb"
}
close(Fout"~resMM.pdb")
close(Fout"~resMM_COU.pdb")
close(Fout"~resMM_VDW.pdb")
close(Fout"~resPBSA.pdb")
close(Fout"~resPBSA_PB.pdb")
close(Fout"~resPBSA_SA.pdb")
close(Fout"~res_MMPBSA.pdb")
}
print "#mol res MM-PBSA(with DH ) MM(with DH ) PBSA " \
"COU(with DH ) VDW PB SA" > pid"~res.dat"
fmt="%-12s %9.3f(%9.3f) %9.3f(%9.3f) %9.3f %9.3f(%9.3f) %9.3f %9.3f %9.3f\n"
for(i=1; i<=Nres; i++) {
tag=Tres[i]
VDWres[tag] /= Nfrm
PBres[tag] /= Nfrm
SAres[tag] /= Nfrm
COUres[tag,0] /= Nfrm
COUres[tag,1] /= Nfrm
MMres[tag,0]=COUres[tag,0]+VDWres[tag]
MMres[tag,1]=COUres[tag,1]+VDWres[tag]
dHres[tag,0]=MMres[tag,0]+PBres[tag]+SAres[tag]
dHres[tag,1]=MMres[tag,1]+PBres[tag]+SAres[tag]
txt=tag; sub(/L~0+/, "Lig ", txt); sub(/P~0+/, "Pro ", txt)
printf fmt, txt,
dHres[tag,0], dHres[tag,1],
MMres[tag,0], MMres[tag,1],
PBres[tag]+SAres[tag],
COUres[tag,0], COUres[tag,1], VDWres[tag], \
PBres[tag], SAres[tag] > pid"~res.dat"
ii=""; if(i==1) ii=sprintf("---------------------------------------\n%-9s", "mean")
txt=""; if(i==Nres) txt="\n"
if(withLig) {
printf "%s%9.3f(%9.3f) %s", ii, dHres[tag,0], dHres[tag,1] , txt > pid"~res_MMPBSA.dat"
printf "%s%9.3f(%9.3f) %s", ii, MMres[tag,0], MMres[tag,1] , txt > pid"~resMM.dat"
printf "%s%9.3f %s", ii, VDWres[tag] , txt > pid"~resMM_VDW.dat"
printf "%s%9.3f(%9.3f) %s", ii, COUres[tag,0], COUres[tag,1], txt > pid"~resMM_COU.dat"
}
printf "%s%9.3f %s", ii, preK*(PBres[tag]+SAres[tag]), txt > pid"~resPBSA.dat"
printf "%s%9.3f %s", ii, preK*(PBres[tag]) , txt > pid"~resPBSA_PB.dat"
printf "%s%9.3f %s", ii, preK*(SAres[tag]) , txt > pid"~resPBSA_SA.dat"
}
fmt="%s%6s%6s\n"
for(i=1; i<=Npro; i++) {
ii=Ipro+i-1
txt=sprintf("%-6s%5d %-4s %3s A%4d %8.3f%8.3f%8.3f", \
"ATOM", ii, type[ii], res[ii], resID[ii], x[ii], y[ii], z[ii])
tag=resPro[i]
if(withLig) {
printf fmt, txt, fixfmt(dHres[tag,0],6), fixfmt(dHres[tag,1],6) > pid"~res_MMPBSA.pdb"
printf fmt, txt, fixfmt(MMres[tag,0]+VDWres[tag],6), fixfmt(MMres[tag,1]+VDWres[tag],6) > pid"~resMM.pdb"
printf fmt, txt, fixfmt(COUres[tag,0],6), fixfmt(COUres[tag,1],6) > pid"~resMM_COU.pdb"
printf fmt, txt, fixfmt(VDWres[tag],6), fixfmt(VDWres[tag],6) > pid"~resMM_VDW.pdb"
}
printf fmt, txt, fixfmt(preK*(PBres[tag]+SAres[tag]),6), fixfmt(preK*(PBres[tag]+SAres[tag]),6) > pid"~resPBSA.pdb"
printf fmt, txt, fixfmt(preK*PBres[tag]+SAres[tag],6), fixfmt(preK*PBres[tag]+SAres[tag],6) > pid"~resPBSA_PB.pdb"
printf fmt, txt, fixfmt(preK*SAres[tag],6), fixfmt(preK*SAres[tag],6) > pid"~resPBSA_SA.pdb"
}
for(i=1; i<=Nlig; i++) {
ii=Ilig+i-1
txt=sprintf("%-6s%5d %-4s %3s A%4d %8.3f%8.3f%8.3f", \
"ATOM", ii, type[ii], res[ii], resID[ii], x[ii], y[ii], z[ii])
tag=resLig[i]
printf fmt, txt, fixfmt(dHres[tag,0],6), fixfmt(dHres[tag,1],6) > pid"~res_MMPBSA.pdb"
printf fmt, txt, fixfmt(MMres[tag,0]+VDWres[tag],6), fixfmt(MMres[tag,1]+VDWres[tag],6) > pid"~resMM.pdb"
printf fmt, txt, fixfmt(COUres[tag,0],6), fixfmt(COUres[tag,1],6) > pid"~resMM_COU.pdb"
printf fmt, txt, fixfmt(VDWres[tag],6), fixfmt(VDWres[tag],6) > pid"~resMM_VDW.pdb"
printf fmt, txt, fixfmt(preK*(PBres[tag]+SAres[tag]),6), fixfmt(preK*(PBres[tag]+SAres[tag]),6) > pid"~resPBSA.pdb"
printf fmt, txt, fixfmt(preK*PBres[tag]+SAres[tag],6), fixfmt(preK*PBres[tag]+SAres[tag],6) > pid"~resPBSA_PB.pdb"
printf fmt, txt, fixfmt(preK*SAres[tag],6), fixfmt(preK*SAres[tag],6) > pid"~resPBSA_SA.pdb"
}
dH[0]=0; dH[1]=0
MM[0]=0; MM[1]=0
COU[0]=0; COU[1]=0
VDW=0; PB=0; SA=0
for(i=1; i<=Nfrm; i++) {
dH[0] += dh[i,0]/Nfrm; dH[1] += dh[i,1]/Nfrm
MM[0] += mm[i,0]/Nfrm; MM[1] += mm[i,1]/Nfrm
COU[0]+= cou[i,0]/Nfrm; COU[1]+= cou[i,1]/Nfrm
VDW += vdw[i]/Nfrm
PB += pb[i]/Nfrm
SA += sa[i]/Nfrm
}
TdS[0]=0; TdS[1]=0 # 1 for DH
if(withLig) {
for(i=1; i<=Nfrm; i++) {
TdS[0] += exp((mm[i,0]-MM[0])/RT2kJ)/Nfrm
TdS[1] += exp((mm[i,1]-MM[1])/RT2kJ)/Nfrm
}
}
TdS[0]=-RT2kJ*log(TdS[0]); TdS[1]=-RT2kJ*log(TdS[1]);
dG[0]=dH[0]-TdS[0]; dG[1]=dH[1]-TdS[1];
Ki[0]=exp(dG[0]/RT2kJ); Ki[1]=exp(dG[1]/RT2kJ)
printf "----------------------------------" \
"|------------------------------------------" \
"|--------------------------------|\n" \
"%-12s %9.3f(%9.3f) | %9.3f(%9.3f) %9.3f %9.3f | %9.3f(%9.3f) %9.3f |\n" \
"%-12s %9.3f(%9.3f) |\n----------------------------------|\n" \
"%-12s %9.3f(%9.3f) kJ/mol = %9.3f(%9.3f) kcal/mol\n" \
"%-12s %9.3E(%9.3E) uM = %9.3E(%9.3E) nM\n" ,
" dH=", dH[0], dH[1], MM[0], MM[1], PB, SA, COU[0], COU[1], VDW,
"-TdS=", -TdS[0], -TdS[1],
" dG=", dG[0], dG[1], dG[0]/4.184, dG[1]/4.184,
" Ki=", Ki[0]*1E6, Ki[1]*1E6, Ki[0]*1E9, Ki[1]*1E9 >> pid"~MMPBSA.dat"
print "+==============================================================+\n"\
" dG = RTln(Ki) = -RTln(KA) \n" \
" Ki = 1/KA = exp(dG/RT) = IC50 = EC50 \n" \
"+==============================================================+\n" \
"| M(mol/L) | m/u/pM | dG(kcal/mol) | dG(kJ/mol) | Affinity |\n" \
"|==============================================================|\n" \
"| 0.1 | 100 mM | -1.364 | -5.708 | |\n" \
"| 0.01 | 10 mM | -2.728 | -11.416 | |\n" \
"| 0.001 | 1 mM | -4.093 | -17.124 | |\n" \
"| 0.0001 | 100 uM | -5.457 | -22.832 | Weak |\n" \
"|--------------------------------------------------------------|\n" \
"| 0.00001 | 10 uM | -6.821 | -28.540 | |\n" \
"| 1.0E-06 | 1 uM | -8.185 | -34.248 | Medium |\n" \
"|--------------------------------------------------------------|\n" \
"| 1.0E-07 | 100 nM | -9.550 | -39.956 | |\n" \
"| 1.0E-08 | 10 nM | -10.914 | -45.664 | |\n" \
"| 1.0E-09 | 1 nM | -12.278 | -51.372 | Strong |\n" \
"|--------------------------------------------------------------|\n" \
"| 1.0E-10 | 100 pM | -13.642 | -57.080 | |\n" \
"| 1.0E-11 | 10 pM | -15.007 | -62.788 | |\n" \
"| 1.0E-12 | 1 pM | -16.371 | -68.496 | Very strong |\n" \
"+==============================================================+\n" >> pid"~MMPBSA.dat"
}
function dimAPBS(file, Imol, minX, maxX, minY, maxY, minZ, maxZ) {
lenX=max(maxX-minX, 0.1); cntX=(maxX+minX)/2
lenY=max(maxY-minY, 0.1); cntY=(maxY+minY)/2
lenZ=max(maxZ-minZ, 0.1); cntZ=(maxZ+minZ)/2
cX =lenX*cfac; fX =min(cX, lenX+fadd)
cY =lenY*cfac; fY =min(cY, lenY+fadd)
cZ =lenZ*cfac; fZ =min(cZ, lenZ+fadd)
levN=4 # 划分级别
t=2^(levN+1)
nX=round(fX/df)-1; nX=max(t*round(nX/t)+1, 33)
nY=round(fY/df)-1; nY=max(t*round(nY/t)+1, 33)
nZ=round(fZ/df)-1; nZ=max(t*round(nZ/t)+1, 33)
if(gridType==0) { # GMXPBSA method
fpre=1; cfac=1.7
fX=lenX+2*fadd; cX=fX*cfac; nX=t*(int(fX/(t*df))+1+fpre)+1
fY=lenY+2*fadd; cY=fY*cfac; nY=t*(int(fY/(t*df))+1+fpre)+1
fZ=lenZ+2*fadd; cZ=fZ*cfac; nZ=t*(int(fZ/(t*df))+1+fpre)+1
}
MGset="mg-auto"
mem = 200*nX*nY*nZ/1024./1024. # MB
# npX=nX; npY=nY; npZ=nZ
# gmem=4000
# ofrac=0.1
# if(mem>=gmem) {
# while(mem>gmem) {
# maxN=max(npX, max(npY, npZ))
# if(maxN==npX) npX = t*((npX-1)/t-1)+1
# else if(maxN==npY) npY = t*((npY-1)/t-1)+1
# else if(maxN==npZ) npZ = t*((npZ-1)/t-1)+1
# mem = 200*npX*npY*npZ/1024./1024
# }
# t=nX/npX; if(t>1) npX = int(t*(1+2*ofrac) + 1.0);
# t=nY/npY; if(t>1) npY = int(t*(1+2*ofrac) + 1.0);
# t=nZ/npZ; if(t>1) npZ = int(t*(1+2*ofrac) + 1.0);
# MGset="mg-para\n ofrac "ofrac"\n pdime "npX" "npY" "npZ
# }
XYZset=" "MGset \
"\n mol "Imol \
"\n dime "nX" "nY" "nZ" # 格点数目, 所需内存: "mem" MB" \
"\n cglen "cX" "cY" "cZ" # 粗略格点长度" \
"\n fglen "fX" "fY" "fZ" # 细密格点长度" \
"\n fgcent "cntX" "cntY" "cntZ" # 细密格点中心" \
"\n cgcent "cntX" "cntY" "cntZ" # 粗略格点中心"
return \
"ELEC name "file"\n" \
XYZset "\n" \
PBEset "\n" \
"end\n\n" \
"ELEC name "file"~VAC\n" \
XYZset "\n" \
PBEset0 "\n" \
"end\n\n" \
"APOLAR name "file"~SAS\n" \
" mol "Imol"\n" \
PBAset"\n" \
"end\n\n" \
"print elecEnergy "file" - "file"~VAC end\n" \
"print apolEnergy "file"~SAS end\n\n"
}
function min(x, y) { return xy ? x : y }
function round(x) { return int(x+0.5) }
function fixfmt(x,m, nint,npre,nexp,str){
str=sprintf("%.9f", x)
sub(/^0+/, "", str);
sub(/^-0+/, "-", str);
nint=index(str, ".")-1
if(-10^m