If a pdbfile is provided, normweightmode becomes sstjetormaxtwocomponent automatically. Else, it is tjet.

There is not any algorithmic changes in this version.
However, I made the following changes to make Python
and Docker installations better:
1-I changed base ubuntu from 20.04 to 22.04.
2-dssp 4.0 in Ubuntu 22.04 have a shitty requirement
 of CRYST record in the pdb file. Therefore, I manually
 installed dssp 3.
3-I needed at least Python 3.9 for importlib.resources
  and I made the required changes both in setup.py and
  Dockerfile for that.

Since esgemme calls two R scripts, namely
pred.R and computePred.R, to do the calculations,
I added them as package resource. They are copied
to the folder where ESGEMME is initiated. I'm
trying to remove this kind of path dependencies
to make the program automatically installable and
portable.

I adapted my code and Dockerfile to accomodate
pip installable version of ESGEMME.

When a protein is bigger than 3276 aa's,
matplotlib can not plot it due to 2^16
horizontal pixel size limit. I changed
plotGEMMEmatrix function and the way I
call it to solve this problem. Now, the
program will divide big proteins into
500 aa chuncks and plot it in that way.

--verbose option controls if epistatic and independent
results files will be written or not. It is False by
default and therefore, the files are not written.

Convert trace->tjet, maxhalftracepchalftracecvhalfcvpc->max
and ssjetormax->sstjetormax in ESGEMME.