Phenome-Mapper

Mapping the genetic basis underlying human phenotypic diversity is a central goal of genomics research. Phenome Mapper offers a comprehensive solution for association analysis between diverse human traits and multiple graph-represented variation loci. Detailed step-by-step workflow to find associations between multiple graph variation loci (such as SV/VNTR) and (sub)phenotypes:

1)Build a sequence graph: Use pangenome graph construction tools (e.g., minigraph, vg, APAV toolkit) to integrate all samples, capturing all structural variations and VNTRs. Step 1: Build a Sequence Graph Goal: Integrate all known and sample-specific genome variations into a comprehensive graph. Tools: minigraph, vg, pggb, or similar. Input: Reference genome (FASTA), sample assemblies/reads, population VCFs. Output: Sequence graph (usually GFA/VG format) capturing SVs, VNTRs, and allelic paths. Example: asuming you have created your work space, installed minigraph, spades for assemblies and dowonloaded reference fasta files.

    bash
        spades.py -1 sample_R1.fastq -2 sample_R2.fastq -o spades_output
            Output: Assembled contigs in spades_output/contigs.fasta
        quast.py spades_output/contigs.fasta -o quast_report/
            # Optionally add: -r reference.fa for reference-based metricsquast flye_assembly/assembly.fasta -o quast_report
            # use your fasta contigs as input to minigraph
        minigraph -x asm -t8 -l -o graph.gfa reference.fa spades_output/contigs.fasta
            # can add more assemblies fasta files minigraph -x asm -t8 -l -o graph.gfa reference.fa sample1.fa sample2.fa ...
            # Or: vg construct -r ref.fa -v cohort.vcf.gz > graph.vg

Index the Sequence Graph for Mapping minigraph outputs a GFA that may require conversion for some mapping/analysis pipelines. If using vg, create XG indexes for rapid mapping and variant calling.

    bash
            # Convert GFA to vg format (if needed)
        vg convert -g graph.gfa > graph.vg
            # Index for Giraffe (vg)
        vg index -x graph.xg -g graph.gg graph.vg

2)Align reads or assemblies: Map sequence data back to the constructed graph using graph-aware aligners (GraphAligner, vg giraffe, Paragraph).

   bash
        vg giraffe -x graph.xg -g graph.gg -m graph.min -d graph.dist \
        -f sample_R1.fastq -f sample_R2.fastq > sample.gam

Call Variants and Genotypes at Graph Loci Generate genotype calls or repeat counts for every sample at each locus in the graph. Tools:(vg pack and vg call) Example:

   bash
       vg pack -x graph.xg -g sample.gam -o sample.pack
       vg call graph.xg -k sample.pack > sample.vcf

3)Create sample-by-locus matrix: For each sample, list allelic state or genotype at each graph variation locus. Build Sample-by-Locus Genotype Matrix Convert your VCF(s) for all samples into a matrix or table, with rows as samples and columns as SV/VNTR loci (with genotype state, repeat lengths, etc.). Tools:(bcftools, vcftools, pandas, or R.) Prepare/Harmonize Phenotype and Covariate Data Ensure your phenotype table matches sample IDs. Include relevant covariates (age, sex, batch effects, ancestry PCs).

```
  bash
    bcftools query -f '%CHROM\t%POS\t%ID[\t%GT]\n' input.vcf.gz > genotype_matrix.tsv  #single sample/multisample)
 ```

#Explanation:
              %CHROM, %POS, %ID: Outputs chromosome, position, and variant ID as locus identifiers.
              [\t%GT]: For each sample, outputs the genotype (0/0, 0/1, etc.), separated by tabs.
              Each line is a locus (SV/VNTR/SNP), and columns after the first three are per sample.
              The output is a tab-delimited file ideal for pandas, R, or spreadsheet analysis.

To include SV or VNTR-specific information (e.g., repeat lengths, SV type):

```
  bash
    bcftools query -f '%CHROM\t%POS\t%ID\t%INFO/SVLEN\t%INFO/REPTYPE[\t%GT]\n' input.vcf.gz > genotype_matrix_with_info.tsv
```    

4)**Annotate associated loci:**For loci showing significant association, examine overlap with genes, regulatory regions, and known disease markers

**Future direction ** Perform Association Analysis Test for statistical association (linear regression/ linear mixed model) between each graph locus (SV/VNTR genotype) and phenotypes/subphenotypes. Tools: plink2, REGENIE, SAIGE, rvtests, R, or Python analysis packages.