small lambda value, PATH CV

[Julia Belyaeva]

Dear plumed users,

I have some uncertainties with my path CV construction.
I would like to accurately study the path of protein N-termini behaviour. Basing on short metaD, an assumption path of the movement of interest with necessary 'start' and 'finish' states was obtained.

I described it with four interatomic distances and made the 'pre-path.pdb' file (40 frames, picture of this distances changes during frames is in attachments):
REMARK ARG=d1,d2,d3,d4 d1=0.316311 d2=0.477923 d3=1.005718 d4=1.610434
END
REMARK ARG=d1,d2,d3,d4 d1=0.315329 d2=0.446196 d3=0.958564 d4=1.561423
...
REMARK ARG=d1,d2,d3,d4 d1=1.607309 d2=1.048764 d3=0.551095 d4=0.313231
END
REMARK ARG=d1,d2,d3,d4 d1=1.511813 d2=1.00816 d3=0.546436 d4=0.362507

END

This file was given to pathtools (metric: EUCLIDEAN), the output was 'final_path.pdb', which I gave to ADAPTIVE_PATH metaD simulation (short, 100 ns):
d1: DISTANCE ATOMS=285,3003
d2: DISTANCE ATOMS=285,2950
d3: DISTANCE ATOMS=285,2924
d4: DISTANCE ATOMS=285,282

p: ADAPTIVE_PATH REFERENCE=final_path.pdb TYPE=EUCLIDEAN WFILE=out_path.pdb UPDATE=1000 WSTRIDE=1000 FIXED=1,30

METAD ...
  LABEL=meta
  ARG=p.gspath
  SIGMA=0.05
  HEIGHT=0.5
  PACE=2500
  GRID_MIN=-1
  GRID_MAX=1.5
  GRID_SPACING=0.01
  CALC_RCT
  ... METAD

uw: UPPER_WALLS ARG=p.gzpath AT=0.6 KAPPA=1000
lw: LOWER_WALLS ARG=p.gzpath AT=-0.6 KAPPA=1000

uws: UPPER_WALLS ARG=p.gspath AT=0.8 KAPPA=1000

lws: LOWER_WALLS ARG=p.gspath AT=-0.6 KAPPA=1000

This output was given to pathtools. And the resulting lambda value was really small, like 0.032.

The question is: what should I do to improve my PCV? And is it possible to use this small lambda value for PATH CV metaD?

Thank you all for your time,
I appreciate your help