data_dir (ver.2.0.0) has 7 sub-directories. 

Each contains:

1. clnvr
   Indel entries of ClinVar database (https://www.ncbi.nlm.nih.gov/clinvar)
   are complied in compressed VCF format (.vcf.gz) with its tabixed file (.vcf.gz.tbi)

2. cosmic
   Indel entries of COSMIC database
   are compiled in compressed VCF format (.vcf.gz) with its tabixed file (.vcf.gz.tbi)

3. dbsnp 
   Indel entries of dbSNP database (https://www.ncbi.nlm.nih.gov/snp)
   are complied in compressed VCF (.vcf.gz) format with its tabixed file (.vcf.gz.tbi)

4. models
   Trained random forest models are pickled (.pkl.gz). 

5. non_somatic
   Known recurrent germline and artifact indels are collected in VCF (.vcf.gz)
   format with its tabixed file (.vcf.gz.tbi)

6. protein
   Conserved domain location (in amino acid position) for each RefSeq isoform.

7. refgene 
   Annotation file in compressed BED format (refCodingExon.bed.gz) 
   and its index file (refCodingExon.bed.gz.tbi) are compiled from refGene
   (http://hgdownload.soe.ucsc.edu/goldenPath/hg38/database/refGene.txt.gz). 
   
   Specification of refCodingExon.bed
    
   Suppose the following isoform 'NM_777' of gene 'AGENE'
   has 5 coding exons (each coding exon denoted with 6 Es (EEEEEE)).
   The genome and cds (coding sequence) are coordinated: 


    genome_pos         90   111  123  145  155  170
                       |    |    |    |    |    | 
               EEEEEE--EEEEEE----EEEEEE----EEEEEE---EEEEEE
               |    |  |    |    |    |    |    |   |    |
    cds_pos    1    77 78   99   100  122  123  138 139  150
             
   
   col1  col2 col3 col4                     col5 col6   col7    
   chrN  123  145  NM_777|AGENE|3|5|100|150  +   90|111 155|170


   column 1 (str): chr1-22, chrX, chrY, chrM
   column 2 (int): start (1-based) of current coding exon
   column 3 (int): end (1-baed) of current coding exon
   column 4 (str|str|int|int|int|int): 
   
             RefSeq_accession|
             GeneSymbol|
             ordinal_of_current_coding_exon|
             total_number_of_coding_exons|
             start_pos_of_cds_in_this_exon|
             the_total_length_of_cds

             Example:
         
                NM_777: the isoform is 'NM_777'
                AGENE: the gene is named 'AGENE'
                3: current coding exon is the 3rd coding exon
                9: this isoform has a total of 9 coding exons
                100: the cds starts at 100 in this coding exon
                120: the length of cds in this isoform is 120
                         
   column 5 (str): strand (+ or -)
   column 6 (int|int): start|end pos of previous coding exon
                       if current exon is the first exon, -1|-1
   column 7 (int|int): start|end pos of next coding exon 
                       if current exon is the last coding exon, -1|-1
