Molecular dynamics (= MD ) is useless, too time-consuming to simulate molecular or protein motion.

(Fig.2)  MD pseudo-potential or force field U's parameters must be artificially chosen.

MD's pseudo-potential = force field with bad parameters often gives wrong results.

This MD's force field potential's parameters must be artificially determined based on experimental results, empirically.

Choosing inappropriate parameters cause wrong results ( this p.5-limitation,  this p.2-left, ), so molecular dynamics (= MD ) unable to predict any physical values or molecular behavior is useless, just impractically time-consuming ( this p.2-1st-paragraph ).

This p.2-2nd-paragraph says
"even small errors in the force field parameters can lead to catastrophic failure in long-time simulation – ML (= machine learning ) force fields may exhibit pathological behavior,.. causing the simulation to become unstable or explode"

Some force field potentials may be adjusted based on quantum mechanical DFT's pseudo-potential (= quantum mechanics or DFT treating the whole molecule as one pseudo-electron model just choosing pseudo-potentials is also empirical, unable to predict any values, so fake ab-initio ), which is also useless too time-consuming and often giving wrong results ( this p.2-left-3rd-paragraph ).

Molecular dynamics (= MD ) has to update positions of all shapeless atoms (inside a molecule) individually (= instead of moving a rigid molecule as a real rigid object with shape ), little by little, based on pseudo-forces caused by differentiating the force-field potential U ( this p.17~36 ) at short time intervals (= each time step must be shorter than 2 fs = 2 ×: 10^-15 s ) repeatedly, many, many times ( this or this p.5-19,  this p.1-14,  this p.2-4,  this p.2-3 ), which takes unrealistically much time, hampering science.

Molecular dynamics (= MD ) takes days ~ months to simulate just μs motion of a protein, so MD cannot simulate much longer (= seconds ~ hours ) important biological reactions.

(Fig.3)  The time-consuming molecular dynamics cannot simulate seconds ~ hours enzymatic reactions or protein's folding.  ← Drug discovery is impossible.

MD is too time-consuming to simulate actual protein's conformational change, folding, so useless.

In molecular dynamics (= MD ) whose one time step (= time interval Δt ) must be less than 2fs (= 10^-15 s,  this p.5 ), simulation of just 1 μs (= 10^-6 s ) molecular motion needs to calculate as many as 10⁹ time steps repeatedly ( 1μs/1fs = 10⁹ ), which takes impractically too much time (= more than days,  this 3rd~4th-paragraphs ).

This-lower current limitation says
"one-microsecond simulation of a relatively small system (estimated to be 25,000 atoms) running on 24 processors, which can take several months to complete ( this p.3-right-1st-paragraph )."

In most molecules or proteins, the time-consuming MD cannot simulate more than 1 μs motion.  ← Even just nano-second (= ns ) protein simulation takes too much time ( this-lower,  this-p.13 table.5 ).

So today's only molecular simulating method = MD cannot simulate important biological (enzymatic) reactions, proteins' folding ( this p.1-right-1st-paragraph ), synthesis, which biological processes happen in milliseconds ~ minutes ~ hours ( this p.2-left-last~right ) time scale.

This p.2-left-1st-paragraph says
"folding times range from microseconds to tens of minutes.... simulating the folding of these larger slower folding proteins using standard molecular dynamics (MD) simulations will remain out of reach for the foreseeable future"

This p.3-left-2nd-paragraph says
"atomistic MD calculations.. being able to compute few nanoseconds with a personal computer, but far from the millisecond to second timescales of domain motions and allosteric transitions occurring in some enzymes"

The unphysical quantum mechanical shapeless atoms and impractically-time-consuming molecular dynamics (= MD ) discourage scientists from developing useful multi-probe atomic force microscopes, hampering technology.

(Fig.4)  The unphysical quantum mechanical shapeless atoms and extremely-time-consuming MD obstructs nano-technology.

Unphysical quantum mechanics and MD pseudo-atomic model hamper innovation.

This current only molecular simulating method = unrealistically-time-consuming molecular dynamics (= MD ) is the main culprit of hampering today's nano-technology by preventing developing useful multi-probe atomic force microscopes..

To build practical cars, machines, planes or laptops, from parts with shapes, we do Not need to use the impractical time-consuming molecular dynamics, or quantum mechanics which just obstructs science.

So if we treat molecules or atoms as real objects (= parts ) with shapes (= discarding useless quantum mechanics, MD methods ), we can design and build any useful molecular devices by manipulating individual atoms or molecules (= real objects with shapes ) through atomic force microscopes with multiple probe tips, like building useful cars, planes and laptops from parts with shapes.

This current impractical time-consuming molecular dynamics based on shapeless quantum mechanical atoms is unable to simulate motions of atoms or molecules manipulated by the multi-probe atomic force microscopes as real objects with shapes, which is why atomic force microscopes have been deadend, useless with only one probe tip for more than 30 years

(Fig.4')  Unphysical quantum mechanical shapeless atoms, impractical MD model has forced scientists to use the useless atomic force microscope with only one probe tip for 40 years.  ← Our nano-technology is deadend, No progress from the useless single probe tip.

Physicists have intentionally stopped developing useful multi-probe atomic force microscope that will make (useless) quantum mechanics and MD obsolete, unnecessary.

In all researches using today's atomic force (or scanning tunneling ) microscope with still only one (useless) single probe tip, scientists always try to express the atoms and molecules as the unphysical quantum mechanical shapeless atom, one-pseudo-electron DFT and the impractically-time-consuming molecular dynamical model (= MD,  this-p.6-right-upper,  this p.10-8. ) in vain.

↑ These current mainstream quantum mechanical DFT (= density functional theory ), and MD can give only useless pseudo-potentials or force field potentials instead of giving concrete shapes to atoms or molecules ( this p.15-16(or p.13-14),  this p.34(or p.13 ) ).

This p.37(or p.16)-1st-paragraph says
"The timescale limitation of MD simulation precludes reproducing this motion pattern. Instead, the tip moves only forward and the depth is reported"  ← The impractically-time-consuming MD model restricts the researchers' free manipulation of atoms.

Atomic force microscope still has only one useless probe tip.  ← Surprisingly No progress for 40 years

Surprisingly, even the present atomic force (or scanning tunneling ) microscope still has only one useless probe tip, even after 36 years have passed since researchers could first manipulate single atoms by the microscope with only one tip in 1989.  ← Our nano-technology is deadend, No progress from the useless single probe tip.

↑ So the unphysical quantum mechanics and molecular dynamical shapeless atomic model prevents the developing useful nano-technology by preventing developing (useful) atomic force microscopes with multiple probes.

↑ In other words, physicists (and academia ) fear that the useful multi-probe atomic force microscopes freely grabbing single molecules with shapes will make the present (useless) quantum mechanics and MD (= unphysical shapeless atomic model ) obsolete and unnecessary.

So physicists (and academia ) intentionally stop developing useful multi-probe atomic force microscopes that will eventually ditch the useless quantum mechanical or MD model, though they already have the technology.  Today's atomic force microscope with only one probe tip is useless except for publishing papers in academic journals.

(Fig.5) Time-consuming molecular dynamics (= MD ) tends to rely on rough approximation = coarse-grained MD with pseudo-atoms sacrificing atomic-level accuracy.

Rough molecular dynamics (= MD ) approximation = coarse-grained MD is unrealistic and inaccurate.

Coarse-grained molecular dynamics (= CGMD ) approximation treating multiple atoms as one pseudo-atom or fictional beads (= This CGMD also has only pseudo-potential, no atomic shape, this p.2 ) is less time-consuming, but more inaccurate, sacrificing atomic-level resolution ( this-last,  this p.1-right-last-paragraph,  this p.23-5,  this-introduction-3rd-4th-paragraph,  this p.9-limitation ).

This p.2-1st-paragraph says
"Although coarse-grain simulation techniques can be used to somewhat avert the time scale limitation, some of the fine (and perhaps critically important) details of atomic interactions are lost during coarse-grain simulations"

This p.2-left-2nd-paragraph says
"the coarse-grained models lack the atomic details of the interactions"

This p.2-introduction says
"However, owing to the loss of information, capturing atomistic properties through CGMD (= coarse-grained molecular dynamics ) or simply CG simulations is a challenge"

This p.1-intro-1st-paragraph says
"many CG (= coarse-grained ) methods suffer from an inability to correctly predict more than one or two material properties"

Even this inaccurate approximate coarse-grained MD is also impractical, too time-consuming by moving pseudo-atoms little by little by differentiating pseudo-potentials artificially adjusted for CGMD ( this p.5-Fig.2 = simulation of just 1 μs CG molecular model took a day,  this p.3-coase grain molecular dynamics ).

Accelerated molecular dynamics (= MD ) is also too time-consuming and impractical.

Accelerated MD is also too time-consuming to simulate milli-second molecular motion ( this p.20= just 30ns simulation,  this p.6 = just 2500ns simulation,  this p.15 = just 3.6μs simulation ).  ← cannot be used for explaining actual long biological reactions.

Monte-Carlo (= MC ) method just tries to move individual atoms randomly, which often gives inaccurate results ( this p.2-right-last-paragraph ).

This Monte-Carlo tends to take much more time than the time-consuming molecular dynamics to simulate proteins, so useless and inefficient ( this p.2-right-1st-paragraph ).

This p.1-left-3rd-paragraph says
"A simulation of protein bovine pancreatic trypsin inhibitor has shown that MD (= time-consuming molecular dynamics ) can be as much as 10 times more efficient than MC (= Monte-Carlo moving atoms randomly is much more time-consuming, useless after all )"

Ab-initio (or first-principle ) molecular dynamics (= AIMD or FPMD ) including Car-Parrinello molecular dynamics (= CPMD ) based on unphysical quantum mechanical one-pseudo-electron DFT pseudo-potential (= DFT is fake ab-initio unable to predict anything ) is much more time-consuming and impractical than the traditional (= classical ) MD ( this p.2-I.introduction,  this p.2-left-1st-paragraph,  this p.23-middle ), so useless.

This p.19(or p.13)-3rd-paragraph says
"several cases are known DFT fails.  the computational demands of CPMD (= ab-initio MD ) are very high.. limiting the length of the simulation to (only) 10-100 picoseconds (= ps,  this p.5-right )
Exchange around polyvalant transition metal ions, takes place on a much longer timescale and is currently beyond the reach of CPMD (and of classical MD ). "

Ab-initio MD or CPMD is useless, unrealistic, just artificially choosing (unphysical) one-pseudo-electron density (= DFT ) = unphysical plane wave electron's cloud whose coefficient parameters (= lacking real electron picture ) are artificially gradually changed at short time intervals ( this p.41,44 ) by using fictitious electron mass ( this p.20,  this p.2-left-last-paragraph ), instead of moving real electrons (= only nuclei are moved classically, paradoxically,  this p.1-left-3rd-paragraph ).

 

Quantum mechanics unable to give shapes and boundaries to individual atoms led to impractical, time-consuming molecular dynamics (= MD ) based on pseudo-potential or force field.

(Fig.6)  Molecular dynamics (= MD ) with shapeless atoms with No boundaries can Not know when two atoms collide. → Updating each atomic position at short time intervals repeatedly (= to avoid two atoms clashing and overlapping ) takes too much time.

Reason why molecular dynamics (= MD ) using unphysical shapeless atoms takes too much time.

When shapeless atoms collide is unpredictable.  → Each time step must be very short (< 2fs ) to avoid erroneous overlap between two shapeless atoms, and repeating those short time step calculations many, many times takes too much time in MD

Today's only method of simulating motions of molecules and proteins is the impractical molecular dynamics (= MD ) using only artificial pseudo-potential called force fields with No atomic shapes.

In real objects with shapes, we can easily predict the objects' motion and when two objects with shapes collide, like predicting billiard balls' trajectory.

But in the unrealistic shapeless quantum mechanical atoms and MD, the time when two shapeless atoms collide with each other is unpredictable due to lack of individual atomic shape and boundary.

So in this impractical molecular dynamical simulation (= MD ), researchers have to calculate and move each atomic position little by little at extremely short time intervals (= each time step Δt must be less than only 2 fs = 2 femto-seconds = 2 × 10^-15 seconds,  this p.4-2nd-paragraph ) repeatedly many many times ( this or this p.5-p.29 ).

↑ Just simulating 1μs molecular motion must repeat calculations of atomic positions as many as 10⁹ times (= 1 μs/ 1 fs = 1 × 10^-6 s/ 1 × 10^-15 s ), which unrealistically takes too much time, and hampers science.

If they choose each time step longer than 2 fs, it increases the chance of two shapeless quantum mechanical atoms colliding and unrealistically overlapping each other.

↑ Because during longer time step, all atoms can move longer distance at once, which increases the risk of unfavorable atomic collision or overlap (= of two atoms ) causing errors and wrong results called "energy explosion or blow up ( this p.37-1st-paragraph,  this p.27-28,  this 5th-paragraph,  this p.28-1st-paragraph,   )".

This p.10-1st-paragraph says
"Too large a time step can cause a molecular dynamics simulation to become unstable, with the total energy rapidly increasing with time. This behavior is often colloquially termed "exploding" and it is caused by devastating atomic collisions that occur when a large time step propagates the positions of two atoms to be nearly overlapping; the (Pauli) repulsive interactions then create a strong force that propels these atoms apart"

This p.1-lower ~ p.2 says
"if a very small value of time step is used, it will not be efficient since it takes a very long calculation time. However, unfortunately, the MD integration algorithm becomes unstable at large time step size. If too large a time step τ is used, the motion of particles becomes unstable due to the very big truncation error in the integration process, so the total energy of the system may rapidly increase with time.... This behavior is often called exploding and caused by devastating atomic collisions that occur when a large time step.. nearly overlapping"

↑ As a result, the MD simulation based on unphysical quantum mechanical shapeless atoms has to choose very short time step (< 2fs ), and repeat those very short time steps, many, many times, which takes impractically much time, hindering scientific progress.