bugfix: Remove Reads Conversion and Various Bugfixes/Improvements

[koisland]

• Updated Git LFS docs for pulling CHM13 reference.
• Fixed including empty assembly files in concat_asm.
• Removed conversion of reads to fastq. Step originally included to avoid error unable to fix when providing bam files.
  • Example of error: https://github.com/mrvollger/SDA/issues/20
• Changed to only allow one extension for hifi reads.
• Add reference centromeres edge base pair additions to the filename.
  • ex. x.600kbp.fa
• Fixed parsing bug in reference centromeres wildcard name.
• Bumped default memory for align_asm_to_ref step.
• Allow configuring alignment threads for alignment threads for align_asm_to_ref step.
• Fix including ct track wildcard for annotating RepeatMasker satellite annotations.
• Fixed incorrect file pattern glob for aggregating HOR monomer data for clustering centromeres.
• Fixed offset cluster image coordinates.
• Reduced full alpha-satellite region filter threshold for dna-brnn output from 1 Mbp to 30 kbp.
• Added filter check for both p and q-arms in map_cens.py
• Added acro specific check in map_cens.py
• Changed bedtools intersection of alignment bed file to use -loj to not remove contigs not aligning to the p-arm.
• Reduced max_alr_len_thr to 100kbp,
• Increased edge len and percentage alpha-satellite repeat required to be considered partial.
• Fixed skipping first line when parsing dna-brnn output.
• Fixed merging multiple/duplicate read files.
• Fixed duplicate sequences not being removed from the concatenated assembly.