Variant Discovery and Prioritization Pipeline

After demultiplexing and FASTQ generation, the paired-end reads were trimmed using TrimGalore v0.4.4. The paired-end reads were then aligned to hg19 human reference genome with BWA v0.7.15-r1140 (11 (link)). We applied the GATK v3.7 pipeline (12 (link)) for indel realignment, duplicate removal, and performed SNP and INDEL discovery as well as share genotype across both the proband and her father’s samples simultaneously according to GATK Best Practices recommendations (14 (link), 15 (link)). Variants were annotated and filtered with Agilent Technologies Bench Lab NGS v5.0.2. To select putative pathogenic variants, filters were applied as follows (Supplementary Fig. 1, see section on supplementary materials given at the end of this article): (i) variants common to proband and her father were filtered out; (ii) low quality reads were filtered out; (iii) only heterozygous variants were retained; (iv) variants predicted as pathogenic (Supplementary Fig. 1) were retained; (v) variants in exonic regions or near splicing sites were retained; (vi) variants with a frequency ≥0.01 in the general population (using Exac database, release 0.3, 1000 Genomes Phase1 release v3.20101123, ESP6500SI-V2, 1000 Genomes Phase 3 release v5.20130502) were filtered out. Relevant nucleotide variants were validated by Sanger sequencing on Applied Biosystems 3730 platform.