Add profile unzipping and conversion option to PopDel view

Add option '-o' to PopDel view for unzipping existing profiles.
Add option '-x' to PopDel profile and PopDel call to write/read
 uncompressed profiles.
Add option '-s' to PopDel view, for changing the index-region size
 when unzipping a profile.
Add options '-c' and '-p' to automate partitioning of a profile
 into chunks of size 'c' base pairs with padding/overlap of 'p' base pairs.

Various fixes and improvements.

Improve memory management for many samples in PopDel call.
Add option '-e'/'--per-sample-rgid' to resolve conflicting read
 group names in PopDel profile.
Fix "too big offset in window"-crash in PopDel profile.
Fix bug in automatic determination of output profile path in
 PopDel profile.

Remove accidentially added files

Remove obsolete distinction based on "informativeness"

- Remove obsolete distinction based on "informativeness" of a read pair prior to likelihood calculations.
- Fix bug in VCF position calculation if the position of a variant was 1.

Remove obsolete distinction based on "informativeness"

- Remove obsolete distinction based on "informativeness" of a read pair prior to likelihood calculations.
- Fix bug in VCF position calculation if the position of a variant was 1.

Fix bug in processing of output

Improve genotyping and position estimation.

- Improve genotyping. The genotyping model for HET variants no
  longer relies on the binomial distribution.
- **Remove parameter '-q' ('--del-ref-ratio')**, because it is no
  longer used in the new genotyping model.
- Improve resolution of deletion starting point estimation
  to 1 bp.
- Change error message for mismatch between contig names found in
  the BAM-file and those specified by the reference version.

Add option for specifying other reference builds and reduced FPs

- Added option to specify the reference build for some human genomes (for details see [Wiki](https://github.com/kehrlab/PopDel/wiki/03.-Create-profiles-with-popdel-profile#specifying-the-reference-genome)).
- Added default sampling regions for the added genome builds.
- Fixed a bug in the merging of subsequent windows that caused some deletions to be split into multiple calls, generating some false positives. This fix also makes the estimation of the position more robust to outliers in the alignments.
- Fixed links to wiki in the README.

Merge changes to README.md

Fix LAD and DAD values in the FORMAT fields

Previously, the LAD and DAD values for the first window of a deletion were displayed, leading to confusing numbers, seemingly contradicting the given genotypes. Now, the median of the numbers making up the LAD and DAD across all windows of the deletion are given instead.