This study involved 77 patients with clinical symptoms of retinal dystrophies, recruited by several centres: the Medical University of Silesia in Katowice, Genomed S.A. in Warsaw, and the Pomeranian Medical University in Szczecin (Table S1 ), and who were included into the cohort between the years 2016 and 2018. The mean age at the time of referral was 40.9 (SD = 16.7).
Patients were subjected to detailed ophthalmological examinations, which involved visual acuity and colour vision testing, autorefractometry, tonometry, perimetry, and dilated fundus examinations, including fundus autofluorescence, optical coherence tomography (OCT), fluorescein angiography, and ERG/EOG, depending on the clinical symptoms. The referring ophthalmologist selected the type of examination for each patient. Results of ophthalmological examinations for a subgroup of retinitis pigmentosa patients were described by Wiacek et al. [19 (link)].
Genetic testing was carried out in the laboratory of Genomed S.A., Warsaw, Poland. Genomic DNA was isolated from the peripheral blood using a total DNA isolation kit (Blood Mini kit, A&A Biotechnology, Gdańsk, Poland) according to the manufacturer’s protocol or as described in [20 (link)]. Genomic DNA was used as input for constructing an exome-enriched library. Two types of the WES enrichment were used during the study: SureSelectXT Clinical Research Exome or Human Exome V5 (Agilent Technologies, Santa Clara, CA, US) and ClinicalExome (Roche, Basel, Switzerland).
The WES libraries were prepared according to the Agilent or Roche protocols using 100 ng of genomic DNA for each sample.
All the libraries were sequenced using the NextSeq500 or HiSeq4000 (Illumina Inc., San Diego, CA, USA) in the PE150 mode, aiming at the mean target coverage above 100×.
Initial processing of BCL files and demultiplexing was done using the Illumina bcl2fastq. Trimmomatic [21 (link)] was applied to trim raw FASTQ files from adapter sequences and low-quality bases. Read mapping to the GRCh37 (hg19) reference genome was performed using the Burrows–Wheeler Alignment tool [22 (link)]. Duplicate read pairs were removed with MarkDuplicates (Picard tools package [23 ]). Alignment files were further processed in accordance with the Genome Analysis Toolkit v3 best practice pipeline [24 ], and Haplotype Caller was used for variant identification. The quality metrics for the alignment and variants were examined, and all variants within the targeted regions, above the default variant quality threshold, were annotated with Annovar [25 (link)]. Further variant filtration and interpretation involved the Gemini framework [26 (link)], enabling an analysis of multiple samples in the search for rare pathogenic variants in the whole exome, as well as an in-house variant analysis software (BroVar v2), using the ACMG variant classification guidelines [27 (link),28 ].
In the first step, targeted exome sequencing data analysis was carried out to identify pathogenic variants in 267 genes associated with retinal dystrophies (based on the RetNet database [1 ]), and subsequently a search for loss-of-function and likely pathogenic variants of the entire exome sequence was performed for samples with negative results. The result set has been limited to rare variants based on the allele frequencies from 1000 Genomes [29 (link)], as well as newly constructed POLGENOM [30 ] databases. Each rare variant (MAF < 0.01 for 1000 Genomes), with the exception of the known pathogenic variants, was assessed for pathogenicity using multiple in silico predictors, including Alamut ver. 2.9.0 (SOPHiA GENETICS SA, Rolle, Switzerland) with all the incorporated predictors, such as SIFT, Mutation Taster, and Polyphen2. Its presence in variant databases such as ClinVar [31 (link)] and HGMD Professional [32 (link)] and in the in-house database was checked. Finally, the variant frequency was compared to the expected frequency for the considered disease. Segregation was assessed when possible. Pathogenic and likely pathogenic variants were confirmed using Sanger sequencing if diagnostic results were issued or the quality of NGS data was below expected. Mutation Surveyor V 5.0.1 (Softgenetics, State College, PA, USA) was employed for the Sanger data analysis. Copy number variants were searched for using XHMM [33 (link)], and an attempt of copy number variation (CNV) analysis was performed with GermlineCNVCaller from GATKv4 package [34 (link)].
A patient case was considered likely solved if one of the confirmatory variants for an autosomal recessive disease was classified as an uncertain significance variant and genotyping results were consistent with clinical data. If such variants were the only ones identified and there were additional data supporting its pathogenicity, a case was also considered likely solved. A partial re-analysis of data was recently performed using updated information on the pathogenicity of presumably causative variants—mainly ClinVar and HGMD Professional—and including re-mapping selected exome datasets to the GRCh38 (hg38) reference genome.
This study was approved by the Ethics Committee of Medical University of Silesia (resolution no. KNW/0022/KB1/105/13, KB 31/2012, and KB/36/A/2013) and adhered to the tenets of The Declaration of Helsinki. Informed written consent was obtained from all the participants.
Patients were subjected to detailed ophthalmological examinations, which involved visual acuity and colour vision testing, autorefractometry, tonometry, perimetry, and dilated fundus examinations, including fundus autofluorescence, optical coherence tomography (OCT), fluorescein angiography, and ERG/EOG, depending on the clinical symptoms. The referring ophthalmologist selected the type of examination for each patient. Results of ophthalmological examinations for a subgroup of retinitis pigmentosa patients were described by Wiacek et al. [19 (link)].
Genetic testing was carried out in the laboratory of Genomed S.A., Warsaw, Poland. Genomic DNA was isolated from the peripheral blood using a total DNA isolation kit (Blood Mini kit, A&A Biotechnology, Gdańsk, Poland) according to the manufacturer’s protocol or as described in [20 (link)]. Genomic DNA was used as input for constructing an exome-enriched library. Two types of the WES enrichment were used during the study: SureSelectXT Clinical Research Exome or Human Exome V5 (Agilent Technologies, Santa Clara, CA, US) and ClinicalExome (Roche, Basel, Switzerland).
The WES libraries were prepared according to the Agilent or Roche protocols using 100 ng of genomic DNA for each sample.
All the libraries were sequenced using the NextSeq500 or HiSeq4000 (Illumina Inc., San Diego, CA, USA) in the PE150 mode, aiming at the mean target coverage above 100×.
Initial processing of BCL files and demultiplexing was done using the Illumina bcl2fastq. Trimmomatic [21 (link)] was applied to trim raw FASTQ files from adapter sequences and low-quality bases. Read mapping to the GRCh37 (hg19) reference genome was performed using the Burrows–Wheeler Alignment tool [22 (link)]. Duplicate read pairs were removed with MarkDuplicates (Picard tools package [23 ]). Alignment files were further processed in accordance with the Genome Analysis Toolkit v3 best practice pipeline [24 ], and Haplotype Caller was used for variant identification. The quality metrics for the alignment and variants were examined, and all variants within the targeted regions, above the default variant quality threshold, were annotated with Annovar [25 (link)]. Further variant filtration and interpretation involved the Gemini framework [26 (link)], enabling an analysis of multiple samples in the search for rare pathogenic variants in the whole exome, as well as an in-house variant analysis software (BroVar v2), using the ACMG variant classification guidelines [27 (link),28 ].
In the first step, targeted exome sequencing data analysis was carried out to identify pathogenic variants in 267 genes associated with retinal dystrophies (based on the RetNet database [1 ]), and subsequently a search for loss-of-function and likely pathogenic variants of the entire exome sequence was performed for samples with negative results. The result set has been limited to rare variants based on the allele frequencies from 1000 Genomes [29 (link)], as well as newly constructed POLGENOM [30 ] databases. Each rare variant (MAF < 0.01 for 1000 Genomes), with the exception of the known pathogenic variants, was assessed for pathogenicity using multiple in silico predictors, including Alamut ver. 2.9.0 (SOPHiA GENETICS SA, Rolle, Switzerland) with all the incorporated predictors, such as SIFT, Mutation Taster, and Polyphen2. Its presence in variant databases such as ClinVar [31 (link)] and HGMD Professional [32 (link)] and in the in-house database was checked. Finally, the variant frequency was compared to the expected frequency for the considered disease. Segregation was assessed when possible. Pathogenic and likely pathogenic variants were confirmed using Sanger sequencing if diagnostic results were issued or the quality of NGS data was below expected. Mutation Surveyor V 5.0.1 (Softgenetics, State College, PA, USA) was employed for the Sanger data analysis. Copy number variants were searched for using XHMM [33 (link)], and an attempt of copy number variation (CNV) analysis was performed with GermlineCNVCaller from GATKv4 package [34 (link)].
A patient case was considered likely solved if one of the confirmatory variants for an autosomal recessive disease was classified as an uncertain significance variant and genotyping results were consistent with clinical data. If such variants were the only ones identified and there were additional data supporting its pathogenicity, a case was also considered likely solved. A partial re-analysis of data was recently performed using updated information on the pathogenicity of presumably causative variants—mainly ClinVar and HGMD Professional—and including re-mapping selected exome datasets to the GRCh38 (hg38) reference genome.
This study was approved by the Ethics Committee of Medical University of Silesia (resolution no. KNW/0022/KB1/105/13, KB 31/2012, and KB/36/A/2013) and adhered to the tenets of The Declaration of Helsinki. Informed written consent was obtained from all the participants.
