Standardisation

What are the features of the human genome?

• Was publicly funded project that started in 1990 and finished around 2003  

• Objective was to determine the entire euchromatic human genome 

• The region that is accessible via sequencing  

• Has lightly packaged chromatin enriched in genes, and represents 92% of the human genome  

How was the HGP funded?

• Had major funding from NIH and DOE  

• Budgeted at 3 billion dollars but cost around 5 billion  

• Was an international project  

What are the features of the draft genome sequence?

• In 2001 the draft genome sequence was released  

• Wasn't complete but allowed analysis and assessment of the genome  

• Contained 22,300 protein coding genes  

• 25% of the genome with 1% are exonic  

• There were significantly more segmental duplications than expected  

• Include nearly identical, repeated sections of DNA, functional gene copies and pseudogenes  

• There was a large number of short interspersed nuclear elements  

• Alu which represented 15% of the genome  

Who did the human genome represent?

• DNA was collected from around 30 donors  

• Came from blood samples in females and sperm samples in males  

• Only a few DNA samples were processed into DNA libraries to ensure anonymity  

• Much of the sequence that went into the draft genome was from 2 females and 1 male  

• Sequencing was murder harder and due to quality control 70% of the draft sequence came from the single male donor  

• Possibly from Buffalo, New York – was he an average human? 

What are the features of the 2001 genome?

• 90% of euchromatin genome was covered  

• There error rate was around 1/1,000  

• Around 150,000 gaps of regions that hadn't been sequenced  

What are the features of the 2003 genome?

• 2003 First release (hg16) 

• Coverage had progressed to around 99%  

• Coverage at 1/10,000 error rate  

• 400 gaps in the euchromatin genome 

What were the features of the 2009 genome?

• GRCh37 was released in 2009 and contained  around 300 gaps 

• Around 550 genes that contained sequencing errors and couldn't be aligned to the genome reference sequence without opening a small gap  

What were the features of the 2013 genome?

• GRCh38 was released in 2013 

• Had 89 unresolved gaps  

• A few hundred genes still required gap-opening to ensure a perfect alignment between transcript and genome  

What were the features of 2022 genome?

• In 2022 there was the T2T 

• Complete human genome sequence with no gaps  

• Claimed to ave no unresolved gaps or mis-alignments  

Which genome sequence is used?

• Many labs still use GRCh37 as a reference sequence but increasing number using GRCh38 

• Not clear if T2T will be used as MANE is based for GRCh38 and migration could cause issues for some genes  

What are patches?

Patches are additional genome sequences with their own identifier which add information to the assembly without disrupting the coordinates or sequences in the reference  

What are fix patches?

• Designed to correct gaps or sequence errors in the genome assembly  

• Provides additional information about the sequence without altering the numbering of bases  

• Could be a patch sating that there is an absent base  

What are novel patches?

• Designed to provide an alternate structure for a chromosomal region  

• Alternative structures are found for the gene  

What happens to the patches with a new genome?

• When new releases of the genome reference are made fix patches can be incorporated into the new assembly  

• Novel patches are still relevant so are retained as alternative loci which represent alternative versions of the assembly at a locus  

What is the HG18 patch?

• Many genes poorly represented in the reference sequence  

• Many were clinically relevant  

• Solution – gene reference sequences  

What is the HG19 patch?

• 300 gaps in the genome reference sequence  

• 550 genes with indel errors where the length of a sequence of a transcript exon is != length of the aligned genomic exon which were clinically relevant genes  

What is the HG38 patch?

• 89 unresolved gaps as of June 2019 

• A few hundred genes with indel errors – a few are within clinically relevant genes  

What is the reference sequence

The HGVSVNC have decided what comprises a reference sequence which is suitable for reporting genetic variation  

What are the different reference sequences?

• UCSC genome browser is a bioinformatics tool and host data used to build the genome browser  

• EBI is a genome annotation and reference sequence archive and also a genome browser  

• NCBI is a reference sequence archive and genome browser  

What are the features of RefSeq?

• Provides genomic reference sequencing files which is more common for analysis  

• Files that are unambiguously format and the chromosome IDs are uniquely labelled with an accession and version number  

• Provides data-files and web tables supporting interconversion of the IDs  

• Has unambiguous ID to each version of the genomic reference sequence, can identify the build and the reference sequence used  

What are the features of transcript reference sequences?

• Needs to contain functional information  

• Exon boundaries, coding sequence start and end positions and location of functional elements  

• Metadata includes: sequence ID, CDS start and end, translation ID/sequence, publications, features  

• It also contains the actual reference sequence  

• RefSeq provides the metadata changes as well 

What are genes?

Genes are a collection of transcript reference sequences and sometimes functional reference sequences and how they may onto a chromosomal reference sequence 

What are alignment files?

Alignment files – data that describe how transcript sequences align to a genome sequence are provided as GFF  

What is the difference between reference sequences?

• Different reference sequence providers have different sets of transcripts and different alignment algorithms  

• Even transcripts with identical sequences may align differently to a given genome build  

• Alignment data can be viewed using a genome browser  

What are the features of RefSeq annotation?

• Transcript reference sequences and genomic sequences are independent and the sequence may disagree  

• Alignment data for a version of a transcript for a gene can align to GRCh37 and 38  

• In rare cases silent re-annotation occur  

• Exon boundaries shift without changes to the reference sequence  

• Why a genomic variant description should always be provided for variants identified by DNA sequencing  

What are the features of ensembl annotation?

• Transcript and genomic sequences not independent and always match  

• If the alignment of a transcript changes, a new version is created  

• Fewer genes have common transcripts between 37/38 

• Makes it tricker to liftover genome builds but stops silent re-alignment  

• The major annotation is fixed  

• Exon boundaries cannot shift without incrementing the version number  

What are the features of GFF3?

• Plain text file that represents genomic and transcriptomic features  

• Includes gene annotation, alignments, and other features  

• Essential in storing information about the genomic elements in their relationships  

What is the GFF3 header section?

• Provides meta-information about the file 

• Includes source, version and additional comments 

What is the GFF3 feature lines?

• Each line represents a genomic or transcriptomic feature  

• Sequence ID: the identifier of the chromosome or sequence where feature is located  

• Source: the source of the features annotation  

• Feature type: type of feature e.g. gene/exon  

• Start position: the starting position of the feature on the sequence  

• End position: the ending position of the feature on the sequence  

• Score: a numerical score or quality score associated with the feature 

What are the features of the GFF3 strand?

• Strand where the feature is located  

• + for forward, - for reverse and . For unknown 

What are the features of GFF3 phase?

The phase of the feature, which is used for features with coding sequences  

What are the features of GFF3 attributes?

• A semicolon-separated list of key-value pairs providing additional information about the feature  

• ID, Name, Parent and Alias  

What are the features of GFF3 attribute tags?

• Key-value pairs that provide additional information about the features  

• ID, name, parent, alias, note, Dbxref, Ontology_term 

What can GFF3 files contain?

• Can be used to store information about gene structures, including the location of exon, introns, transcripts and other genomic features  

• Critical role in genome annotation and transcriptome analysis  

What is the locus reference genome?

• LGRs provide a stable gene sequence and annotations  

• Contain a minimal set of transcripts that could be used for variant reporting and annotation  

• Include a minimum set of transcripts for reporting at a locus and manually created by expert scientists  

How were LRGs producing gene reference sequences?

Gene reference sequences were produced and the transcripts for each gene were annotated against the reference sequence to be sufficient for clinical reporting  

What are the features of stability?

• Includes transcript sequences that are stable and independent of changes to transcript models  

• Establish a coordinate system that is independent from upgrades to the reference genome assembly and provides mappings to present and past assemblies  

• Use a unique and stable identifier that is not versioned  

How has clinical variant changed over time?

• Part of the effort to rationalise differences in NCBI and EBI gene sets  

• Aim to achieve faster convergence between NCBBI and EBI on key high value annotations to provide a common minimal set of transcripts per gene  

• Facilitate unambiguous multi-directional data exchange between NCBI, EBI and the reference genome assemblies  

• Define the relationship between legacy and community-requested sequences and the reference assembly  

• Ensure compatibility with genome reference assembly-based NGS variant reporting systems  

How is support important?

• Integrated into the Ensembl, NCBI and UCSC genome browsers to allow visualisation in genomic context, with all other existing annotations  

• Compatible with the HGVS nomenclature and supported by nomenclature checker systems  

• Generated and maintained by the NCBI and EBI systems  

What are the features of 3rd party annotations?

• LGRs weren't reference sequences but 3^rd party annotations  

• Any discrepancy between the LTR transcript annotations and the source reference sequence were presented in the metadata  

What are the LGRs and GRCh38?

• LGRs and gene reference sequences are obsolete  

• Many are very old based on 37 or earlier so cannot be updated  

• Some LGRs are refractory to creation of Ensembl reference sequences  

• 3^rd party annotations rather than a true reference sequence archive 

What is the summary of LGRs?

• LGRs were designed as static reference sequences for standardised clinical practice  

• From a bygone era pre-dating good genome refernece sequence  

• Static so difficult to incorporate into new knowledge  

• Cannot be updated without creation of a new LGR  

• Many transcripts are not aligned to 38 as they are too old  

• Do not represent all of the gene's clinically relevant transcripts so can't accurately map  

• LGRs had low coverage and the majority of genes did not have an LGR record  

How are MANE and NIH matched?

• Matched annotation between NIH and EBI (MANE) built on the LGR product  

• Designed to ensure each gene in the genome had at least one identical transcript from end to end, including exon-break points  H

How does MANE replace LGRs?

• LRG was an Ensembl and NIH project  

• LGRS have poor coverage <15% where MANE is at >98%  

• LGRs are static don't allow evolution so out of date  

• Genome 38 is good quality so no longer need independent LRG sequences  

• LGR was discontinued in early 2022  

• LGRs are used to inform MANE  

How have MANEs evolved?

• LGRs have never been independent of reference sequence  

• MANE ensures the ID is entirely interchangeable between RefSeq and Ensembl 

• Unlike LRGs MANE intends to include a comprehensive set of clinical transcripts based on the most up to date scientific and clinical evidence 

How does MANE 3rd party annotations work?

• Could have short hand nomenclature that would be out of context when compared to a MANE transcript or protein  

• Can lead to opportunities for treatment being missed out   

What are the features of stable identifiers?

• Variant reporting can lose context if the reference sequence it is being compared to is lost  

• Stable IDs are used throughout genetics 

• They don't change over time and not mutable  

• Only identify a single stable sequence, both ID and sequence won't change over time  

• If the sequence needs to change the ID also needs to change  

• Can do the same with the gene ID  

What are variants?

Variants refer to sequence-level changes between query sequences in comparison to aligned position within a reference sequence  

What does a reference sequence need to have?

• Widely available  

• Well annotated  

• Where translation is initiated  

• Where the UTRs are  

• Where the exon boundaries are  

• Maintained and updated when necessary  

• Uniquely identifiable  

What does a good reference sequence have?

• Have metadata including  

• Accession numbers and versions, definitions, the organism, related articles, annotations, revision history  

What is a mutation?

A mutation is defined as a permeant change in the nucleotide sequence  W

What is a polymorphism?

• A polymorphism is a variant that has a frequency above 1%  

• Confusing as they assume pathogenic vs benign affects  

What are variants?

Variants are changes between a query sequence and a reference sequence  

What are alleles?

Alleles are a set of specific changes within a reference sequence, one or multiple changes  

What are haplotypes?

Haplotype is a specific combination of alleles that occur together on single reference sequence in one or more individuals  

What is the genotype?

Genotype is a specific combination of haplotypes across multiple reference sequences  

What are the features of nomenclature?

• The sequence ontology is a set of terms and relationships used to describe the features and attributes of biological sequence  

• Sequence ontology includes different kinds of features which can be located on the sequence  

• Biological features are those which are defined by their disposition to be involved in a biological process  

How has HUGO changed over time?

• 1949: variant identified using cellulose acetate electrophoresis  

• Later shown to be due to abnormal b-globin chain  

• 1956: abnormality located by peptide fingerprinting  

• Normal and sickle b-globin chains and digested with protease  

• Fragments were separated by chromatography and electrophoresis  

• Confirmed by 1959 by protein sequencing prior to the genetic code being worked out  

• Variant was Glu to Val and A to T transversion  

Describe the sequence features?

Reference sequence contain an initiation Met at position 1 where more proteins don't  

What would the sequence be at different levels?

What are the features of MANE Plus?

MANE plus transcripts are additional clinically relevant reference transcripts for genes that have alternative splice transcript isoforms  

What are the features of EDS?

• COL5A1 encodes two a1 and an a2 chain for collagen V heterotrimer a1,a1,a2  

• Has two alternatively spliced transcript variants using two mutually exclusive exons which lead to protein isoforms  

• A which is 70% retains exon 64a 

• B which is 30% retains 64b  

What does this mean?

What does this mean?

Why might there be no MANE plus transcript?

There is no MANE Plus Clinical transcript because they didn't know which sequence was clinically relevant