Alamut batch

Illumina adaptors were trimmed using Flexbar (Dodt et al. 2012) with the following settings: minimum 6 bp overlap between adapter and read sequence allowing 2 bp mismatches, trimming of 3′end until base quality of Phred‐score 25, and a minimum read length of 30. Trimmed fastq‐files were aligned to the human reference sequence assembly Feb.2009 GRCh37/hg19 (UCSC Genome Browser) with the Burrows‐Wheeler Aligner (BWA)‐MEM algorithm (Li and Durbin 2009) to generate BAM‐files. The following settings were used with the BWA‐MEM algorithm: Mismatch penalty of 20 and gap penalty of 4. Variant Caller Files (VCFs) were generated using HaplotypeCaller of GATK ver. 2.6.5 (McKenna et al. 2010) with a minimum emission confidence threshold of Phred‐score 10 and a calling confidence threshold of Phred‐score 30. Postvariant analysis on VCFs was carried out in the statistical software R (R core team, version 3.1.1, URL http://www.R-project.org). Variants were accepted if they had a minimum coverage of 25 and heterozygote variants calls were accepted if the read frequency of the minor variant was >0.15. Accepted variants were analyzed using Alamut Batch (Interactive Biosoftware, Rouen, France).

Variants were called using LoFreq (version 2.1.1), FreeBayes (version 0.9.9) and proprietary methods. VariantPlex pipeline parameters were adjusted to accommodate the large footprint of SNiPDx and minimize background noise within variant calls. Variant calls <300 base pairs from the nearest gene-specific primers within regions of interest in reads with a minimum base quality of 22, and with a minimum allele fraction of 0.02, were reported. Genes, transcripts and consequences of variants were accessed through Alamut Batch (Interactive BioSoftware) using database version 1.5-2020.11.25. A variant call format file for each sample was generated using VCFtools version 0.1.11.

Library preparation and capture for exome sequencing was performed at the Kinghorn Centre for Clinical Genomics at the Garvan Institute of Medical Research using Agilent SureSelect Clinical Research Exome v2 (Agilent). Sequencing was performed on a HiSeqX (Illumina) to an average depth of coverage of 100× across captured regions. Alignments and variant calls were generated using SoftGenetics NextGene (version 2.4.1, SoftGenetics) to the February 2009 human genome assembly (GRCh37/hg19), and variant calls were restricted to coding regions and the canonical splice sites of an epileptic encephalopathy gene panel. Variants identified were then annotated using Alamut Batch (version 1.9, Interactive Biosoftware) and classified according to the American College of Medical Genetics (ACMG) criteria (Richards et al. 2015 (link)).

Annotation and ranking of SNVs/indels and structural variations were performed, respectively, by VaRank (Geoffroy et al, 2015) (in combination with Alamut Batch, Interactive Biosoftware, Rouen, France) and by AnnotSV (Geoffroy et al, 2018a). Variant effect on the nearest splice site was predicted using MaxEntScan (Yeo & Burge, 2004), NNSplice (Reese et al, 1997) and Splice Site Finder (Shapiro & Senapathy, 1987). Very stringent filtering criteria were applied to filter out non‐pathogenic variants: (i) variants represented with an allele frequency of more than 1% in public variation databases including the 1000 Genomes (1000 Genomes Project Consortium et al, 2015), the gnomAD database (Lek et al, 2016), the DGV (MacDonald et al, 2014) or our internal exome database, and (ii) variants in 5′ and 3′ UTR, downstream, upstream, intronic and synonymous locations without pathogenic prediction of local splice effect. The PSMC3 nomenclature is based on the accession number NM_002804.4 from the RefSeq database (O'Leary et al, 2016). Genomic coordinates are defined according to GRCh37/hg19 assembly downloaded from the University of California Santa Cruz (UCSC) genome browser (Tyner et al, 2017).

The customized panel for oral–facial–digital (OFD) syndrome included five genes: C2CD3, CPLANE1, DDX59, OFD1, and TCTN3. Baits for the OFD gene panel and whole-exome sequencing (WES), which targeted on comprehensive medical exomes, were purchased from Agilent. Both the targeted sequencing library and the WES library for patient 33 were constructed using the Agilent SureSelect clinical research exome kit, except that the hybrid captures were performed using different baits as described above. Sequencing was conducted on a NextSeq550 sequencer (Illumina) with 100 bp paired-end reads. The DNA sequence was aligned to the human reference genome (UCSC hg19). We built a sequence processing pipeline for analyzing both the targeted gene panel and the WES data by incorporating two commercial software, CLC biomedical genomic workbench from QIAGEN and AlamutBatch from Interactive Biosoftware. The evaluation and the classification of variants were performed following the American College of Medical Genetics (ACMG) recommendations (Richards et al., 2015 (link)).