Precise identification of cancer cells from allelic imbalances in single cell transcriptomes.

A fundamental step of tumour single cell mRNA analysis is separating cancer and non-cancer cells. We show that the common approach to separation, using shifts in average expression, can lead to erroneous biological conclusions. By contrast, allelic imbalances representing copy number changes directly detect the cancer genotype and accurately separate cancer from non-cancer cells. Our findings provide a definitive approach to identifying cancer cells from single cell mRNA sequencing data.

Impact of homologous recombination status and responses with veliparib combined with first-line chemotherapy in ovarian cancer in the Phase 3 VELIA/GOG-3005 study.

OBJECTIVE: In the Phase 3 VELIA trial (NCT02470585), PARP inhibitor (PARPi) veliparib was combined with first-line chemotherapy and continued as maintenance for patients with ovarian carcinoma enrolled regardless of chemotherapy response or biomarker status. Here, we report exploratory analyses of the impact of homologous recombination deficient (HRD) or proficient (HRP) status on progression-free survival (PFS) and objective response rates during chemotherapy. METHODS: Women with Stage III-IV ovarian carcinoma were randomized to veliparib-throughout, veliparib-combination-only, or placebo. Stratification factors included timing of surgery and germline BRCA mutation status. HRD status was dichotomized at genomic instability score 33. During combination therapy, CA-125 levels were measured at baseline and each cycle; radiographic responses were assessed every 9 weeks. RESULTS: Of 1140 patients randomized, 742 had BRCA wild type (BRCAwt) tumors (HRP, n = 373; HRD/BRCAwt, n = 329). PFS hazard ratios between veliparib-throughout versus control were similar in both BRCAwt populations (HRD/BRCAwt: 22.9 vs 19.8 months; hazard ratio 0.76; 95% confidence interval [CI] 0.53-1.09; HRP: 15.0 vs 11.5 months; hazard ratio 0.765; 95% CI 0.56-1.04). By Cycle 3, the proportion with ≥90% CA-125 reduction from baseline was higher in those receiving veliparib (pooled arms) versus control (34% vs 23%; P = 0.0004); particularly in BRCAwt and HRP subgroups. Complete response rates among patients with measurable disease after surgery were 24% with veliparib (pooled arms) and 18% with control. CONCLUSIONS: These results potentially broaden opportunities for PARPi utilization among patients who would not qualify for frontline PARPi maintenance based on other trials.

Chromatin accessibility and gene expression during adipocyte differentiation identify context-dependent effects at cardiometabolic GWAS loci.

Chromatin accessibility and gene expression in relevant cell contexts can guide identification of regulatory elements and mechanisms at genome-wide association study (GWAS) loci. To identify regulatory elements that display differential activity across adipocyte differentiation, we performed ATAC-seq and RNA-seq in a human cell model of preadipocytes and adipocytes at days 4 and 14 of differentiation. For comparison, we created a consensus map of ATAC-seq peaks in 11 human subcutaneous adipose tissue samples. We identified 58,387 context-dependent chromatin accessibility peaks and 3,090 context-dependent genes between all timepoint comparisons (log2 fold change>1, FDR<5%) with 15,919 adipocyte- and 18,244 preadipocyte-dependent peaks. Adipocyte-dependent peaks showed increased overlap (60.1%) with Roadmap Epigenomics adipocyte nuclei enhancers compared to preadipocyte-dependent peaks (11.5%). We linked context-dependent peaks to genes based on adipocyte promoter capture Hi-C data, overlap with adipose eQTL variants, and context-dependent gene expression. Of 16,167 context-dependent peaks linked to a gene, 5,145 were linked by two or more strategies to 1,670 genes. Among GWAS loci for cardiometabolic traits, adipocyte-dependent peaks, but not preadipocyte-dependent peaks, showed significant enrichment (LD score regression P<0.005) for waist-to-hip ratio and modest enrichment (P < 0.05) for HDL-cholesterol. We identified 659 peaks linked to 503 genes by two or more approaches and overlapping a GWAS signal, suggesting a regulatory mechanism at these loci. To identify variants that may alter chromatin accessibility between timepoints, we identified 582 variants in 454 context-dependent peaks that demonstrated allelic imbalance in accessibility (FDR<5%), of which 55 peaks also overlapped GWAS variants. At one GWAS locus for palmitoleic acid, rs603424 was located in an adipocyte-dependent peak linked to SCD and exhibited allelic differences in transcriptional activity in adipocytes (P = 0.003) but not preadipocytes (P = 0.09). These results demonstrate that context-dependent peaks and genes can guide discovery of regulatory variants at GWAS loci and aid identification of regulatory mechanisms.

Malignant transformation of salivary gland pleomorphic adenoma: proof of principle.

Supposed risk of malignant transformation of salivary gland pleomorphic adenoma (SGPA) is an important reason for aggressive retreatment in recurrent pleomorphic adenoma (RPA). However, although the diagnostic category 'carcinoma ex-pleomorphic adenoma' suggests that malignant transformation of a pleomorphic adenoma is a regular event, this has to date not been shown to occur in sequential lesions of one patient. Here, we show the molecular events in transformation to malignancy of a pleomorphic adenoma of the parotid gland. Detailed molecular analysis revealed an LIFR/PLAG1 translocation characteristic for pleomorphic adenoma and, next to this, a PIK3R1 frameshift mutation and several allelic imbalances. In subsequent malignant recurrences, the same LIFR/PLAG1 translocation, PIK3R1 frameshift mutation, and allelic imbalances were present in addition to TP53 mutations. Thus, this case not only shows malignant transformation of SGPA, but also demonstrates that molecular analysis can be of help in recognising malignancy in the rare instance of RPA.

Transcriptional bursts explain autosomal random monoallelic expression and affect allelic imbalance.

Transcriptional bursts render substantial biological noise in cellular transcriptomes. Here, we investigated the theoretical extent of allelic expression resulting from transcriptional bursting and how it compared to the amount biallelic, monoallelic and allele-biased expression observed in single-cell RNA-sequencing (scRNA-seq) data. We found that transcriptional bursting can explain the allelic expression patterns observed in single cells, including the frequent observations of autosomal monoallelic gene expression. Importantly, we identified that the burst frequency largely determined the fraction of cells with monoallelic expression, whereas the burst size had little effect on monoallelic observations. The high consistency between the bursting model predictions and scRNA-seq observations made it possible to assess the heterogeneity of a group of cells as their deviation in allelic observations from the expected. Finally, both burst frequency and size contributed to allelic imbalance observations and reinforced that studies of allelic imbalance can be confounded from the inherent noise in transcriptional bursting. Altogether, we demonstrate that allele-level transcriptional bursting renders widespread, although predictable, amounts of monoallelic and biallelic expression in single cells and cell populations.

Homologous recombination deficiency (HRD) score in germline BRCA2- versus ATM-altered prostate cancer.

The homologous recombination deficiency (HRD) score integrates three DNA-based measures of genomic instability, and has been understudied in prostate cancer. Given the recent FDA approval of two PARP inhibitors for prostate cancer, HRD score analysis could help to refine treatment selection. We assessed HRD score (defined as the sum of loss-of-heterozygosity, telomeric allelic imbalance, and large-scale state transitions) in three cohorts of primary prostate cancer, including a Johns Hopkins University (JHU) cohort with germline mutations in BRCA2, ATM, or CHEK2 (n = 64), the TCGA cohort (n = 391), and the PROGENE cohort (n = 102). In the JHU cohort, tumors with germline BRCA2 mutations had higher HRD scores (median = 27) than those with germline ATM or CHEK2 mutations (median = 16.5 [p = 0.029] and 9 [p < 0.001], respectively). For TCGA tumors without underlying HR pathway mutations, the median HRD score was 11, significantly lower than ovarian carcinoma lacking BRCA1/2 mutations (median = 28). In the absence of HR gene mutations, the median HRD score was unexpectedly higher among prostate cancers with TP53 mutations versus those without (17 vs. 11; p = 0.015); this finding was confirmed in the PROGENE cohort (24 vs. 16; p = 0.001). Finally, among eight BRCA2-altered patients who received olaparib, progression-free survival trended longer in those with HRD scores above versus below the median (14.9 vs. 9.9 months). We conclude that HRD scores are low in primary prostate cancer and higher in cases with germline BRCA2 or somatic TP53 mutations. Germline BRCA2-altered cases have significantly higher HRD scores than germline ATM-altered or CHEK2-altered cases, consistent with the lower efficacy of PARP inhibitors among the latter.

Molecular alterations in gastric cancer and the surrounding intestinal metaplastic mucosa: an analysis of isolated glands.

BACKGROUND: Intestinal metaplasias (IMs) are generally regarded as pre-neoplastic gastric lesions. However, molecular alterations including genetic and epigenetic changes occurring in individual IM glands are not well defined. AIMS: We sought to identify DNA methylation status, microsatellite instability (MSI) and allelic imbalance (AI) occurring in individual IM glands and non-IM glands within the same mucosa. METHODS: We divided examined isolated gland obtained from GC into 4 components: isolated cancer, antral isolated intestinal metaplastic tissue, antral isolated non-metaplastic gland and isolated non-metaplastic gland derived from the greater curvature of the most distant gastric body without mucosal atrophy. We examined AI and microsatellite instability statuses using PCR-based microsatellite analysis. Next, the DNA methylation status (high methylation epigenome [HME], intermediate methylation epigenome [IME], and low methylation epigenome [LME]) was investigated. DNA methylation analysis of CDKN2A, mir34-b/c and MLHI genes was also performed. RESULTS: Although antral isolated IM glands were characterized by IME, isolated non-IM glands showed LME. In isolated cancer glands, HME was frequently found, compared with isolated non-IM glands. DNA methylation of mir34-b/c was common in isolated cancer and IM glands, whereas DNA methylation of CDKN2A was a rare event in isolated samples. The MLH1 gene was not methylated in isolated non-IM glands. Although multiple AIs were frequently found in isolated cancer glands, a few AIs were detected in isolated IM glands. CONCLUSIONS: We suggest that the DNA methylation status and the status of the mir34-b/c gene among isolated samples of IMs and isolated non-IM glands have an impact on IM development.

Differential 3D chromatin organization and gene activity in genomic imprinting.

Genomic imprinting gives rise to parent-of-origin dependent allelic gene expression. Most imprinted genes cluster in domains where differentially methylated regions (DMRs)-carrying CpG methylation on one parental allele-regulate their activity. Several imprinted DMRs bind CTCF on the non-methylated allele. CTCF structures TADs ('Topologically Associating Domains'), which are structural units of transcriptional regulation. Recent investigations show that imprinted domains are embedded within TADs that are similar on both parental chromosomes. Within these TADs, however, allelic subdomains are structured by combinations of mono-allelic and bi-allelic CTCF binding that guide imprinted expression. This emerging view indicates that imprinted chromosomal domains should be considered at the overarching TAD level, and questions how CTCF integrates with other regulatory proteins and lncRNAs to achieve imprinted transcriptional programs.

MeCP2 is involved in random mono-allelic expression for a subset of human autosomal genes.

Widespread random monoallelic gene expression (RMAE) effects influence about 10% of human genes. However, the mechanisms by which RME of autosomal genes is established and those by which it is maintained both remain open questions. Because the choice of allelic expression is randomly performed cell-by-cell, the RMAE mechanism is not observable in non-clonal cell populations or in whole tissues. Several target genes of MeCP2, the gene involved in Rett syndrome (RTT), have been previously described as subject to RMAE, suggesting that MeCP2 may be involved in the establishment and/or maintenance of RME of autosomal genes. To improve our knowledge on this largely unknown phenomenon, and to study the role of MeCP2 in RMAE, we compared RMA gene expression profiles in clonal cell cultures expressing wild-type MeCP2 versus mutant MeCP2 from a RTT patient carrying a pathogenic non-sense variant. Our data clearly demonstrated that MeCP2 deficiency does not affect significantly allelic gene expression of X-linked genes, imprinted genes as well as the RMAE profile in the majority of genes. However, the functional deficiency in MeCP2 appeared to disrupt the mono-allelic or the bi-allelic expression of at least 49 genes allowing us to define a specific signature of MECP2 mutated clones.

Allelic Imbalance of Recurrently Mutated Genes in Acute Myeloid Leukaemia.

The patho-mechanism of somatic driver mutations in cancer usually involves transcription, but the proportion of mutations and wild-type alleles transcribed from DNA to RNA is largely unknown. We systematically compared the variant allele frequencies of recurrently mutated genes in DNA and RNA sequencing data of 246 acute myeloid leukaemia (AML) patients. We observed that 95% of all detected variants were transcribed while the rest were not detectable in RNA sequencing with a minimum read-depth cut-off (10x). Our analysis focusing on 11 genes harbouring recurring mutations demonstrated allelic imbalance (AI) in most patients. GATA2, RUNX1, TET2, SRSF2, IDH2, PTPN11, WT1, NPM1 and CEBPA showed significant AIs. While the effect size was small in general, GATA2 exhibited the largest allelic imbalance. By pooling heterogeneous data from three independent AML cohorts with paired DNA and RNA sequencing (N = 253), we could validate the preferential transcription of GATA2-mutated alleles. Differential expression analysis of the genes with significant AI showed no significant differential gene and isoform expression for the mutated genes, between mutated and wild-type patients. In conclusion, our analyses identified AI in nine out of eleven recurrently mutated genes. AI might be a common phenomenon in AML which potentially contributes to leukaemogenesis.

Profiling allele-specific gene expression in brains from individuals with autism spectrum disorder reveals preferential minor allele usage.

One fundamental but understudied mechanism of gene regulation in disease is allele-specific expression (ASE), the preferential expression of one allele. We leveraged RNA-sequencing data from human brain to assess ASE in autism spectrum disorder (ASD). When ASE is observed in ASD, the allele with lower population frequency (minor allele) is preferentially more highly expressed than the major allele, opposite to the canonical pattern. Importantly, genes showing ASE in ASD are enriched in those downregulated in ASD postmortem brains and in genes harboring de novo mutations in ASD. Two regions, 14q32 and 15q11, containing all known orphan C/D box small nucleolar RNAs (snoRNAs), are particularly enriched in shifts to higher minor allele expression. We demonstrate that this allele shifting enhances snoRNA-targeted splicing changes in ASD-related target genes in idiopathic ASD and 15q11-q13 duplication syndrome. Together, these results implicate allelic imbalance and dysregulation of orphan C/D box snoRNAs in ASD pathogenesis.

Comparison of quantitative trait loci methods: Total expression and allelic imbalance method in brain RNA-seq.

BACKGROUND: Of the 108 Schizophrenia (SZ) risk-loci discovered through genome-wide association studies (GWAS), 96 are not altering the sequence of any protein. Evidence linking non-coding risk-SNPs and genes may be established using expression quantitative trait loci (eQTL). However, other approaches such allelic expression quantitative trait loci (aeQTL) also may be of use. METHODS: We applied both the eQTL and aeQTL analysis to a biobank of deeply sequenced RNA from 680 dorso-lateral pre-frontal cortex (DLPFC) samples. For each of 340 genes proximal to the SZ risk-SNPs, we asked how much SNP-genotype affected total expression (eQTL), as well as how much the expression ratio between the two alleles differed from 1:1 as a consequence of the risk-SNP genotype (aeQTL). RESULTS: We analyzed overlap with comparable eQTL-findings: 16 of the 30 risk-SNPs known to have gene-level eQTL also had gene-level aeQTL effects. 6 of 21 risk-SNPs with known splice-eQTL had exon-aeQTL effects. 12 novel potential risk genes were identified with the aeQTL approach, while 55 tested SNP-pairs were found as eQTL but not aeQTL. Of the tested 108 loci we could find at least one gene to be associated with 21 of the risk-SNPs using gene-level aeQTL, and with an additional 18 risk-SNPs using exon-level aeQTL. CONCLUSION: Our results suggest that the aeQTL strategy complements the eQTL approach to susceptibility gene identification.

Homologous recombination deficiency in triple negative breast cancer.

Triple negative breast cancer (TNBC) represents a heterogeneous subtype of breast cancer characterized by an unfavorable prognosis due to its aggressive biology. The median overall survival (OS) for patients with metastatic TNBC is around 9-12 months with conventional cytotoxic agents. Considering this suboptimal outcome, which is induced despite of medical treatment, new therapeutic strategies would be urgently needed. The ultimate goal of precision medicine is to identify specific molecular alterations that permit considering effective targeted drug(s). Germline BRCA mutations occur in 10-20% of TNBC patients while somatic mutations occur in 3-5% of them. Alterations in the homologous recombination (HR) system are typical of BRCA mutant tumors, but can also be identified in tumors that do not carry this mutation, defining a subgroup of patients referred to as BRCAness. In this review, we focus on the role of homologous recombination deficiency (HRD) as both predictive and prognostic factor in different settings of TNBC patients treated with DNA damaging drugs and poly ADP ribose polymerase (PARP) inhibitors.

Survey of allele specific expression in bovine muscle.

Allelic imbalance is a common phenomenon in mammals that plays an important role in gene regulation. An Allele Specific Expression (ASE) approach can be used to detect variants with a cis-regulatory effect on gene expression. In cattle, this type of study has only been done once in Holstein. In our study we performed a genome-wide analysis of ASE in 19 Limousine muscle samples. We identified 5,658 ASE SNPs (Single Nucleotide Polymorphisms showing allele specific expression) in 13% of genes with detectable expression in the Longissimus thoraci muscle. Interestingly we found allelic imbalance in AOX1, PALLD and CAST genes. We also found 2,107 ASE SNPs located within genomic regions associated with meat or carcass traits. In order to identify causative cis-regulatory variants explaining ASE we searched for SNPs altering binding sites of transcription factors or microRNAs. We identified one SNP in the 3'UTR region of PRNP that could be a causal regulatory variant modifying binding sites of several miRNAs. We showed that ASE is frequent within our muscle samples. Our data could be used to elucidate the molecular mechanisms underlying gene expression imbalance.

Allele-specific RNA imaging shows that allelic imbalances can arise in tissues through transcriptional bursting.

Extensive cell-to-cell variation exists even among putatively identical cells, and there is great interest in understanding how the properties of transcription relate to this heterogeneity. Differential expression from the two gene copies in diploid cells could potentially contribute, yet our ability to measure from which gene copy individual RNAs originated remains limited, particularly in the context of tissues. Here, we demonstrate quantitative, single molecule allele-specific RNA FISH adapted for use on tissue sections, allowing us to determine the chromosome of origin of individual RNA molecules in formaldehyde-fixed tissues. We used this method to visualize the allele-specific expression of Xist and multiple autosomal genes in mouse kidney. By combining these data with mathematical modeling, we evaluated models for allele-specific heterogeneity, in particular demonstrating that apparent expression from only one of the alleles in single cells can arise as a consequence of low-level mRNA abundance and transcriptional bursting.

Annotating Transcriptional Effects of Genetic Variants in Disease-Relevant Tissue: Transcriptome-Wide Allelic Imbalance in Osteoarthritic Cartilage.

OBJECTIVE: Multiple single-nucleotide polymorphisms (SNPs) conferring susceptibility to osteoarthritis (OA) mark imbalanced expression of positional genes in articular cartilage, reflected by unequally expressed alleles among heterozygotes (allelic imbalance [AI]). We undertook this study to explore the articular cartilage transcriptome from OA patients for AI events to identify putative disease-driving genetic variation. METHODS: AI was assessed in 42 preserved and 5 lesioned OA cartilage samples (from the Research Arthritis and Articular Cartilage study) for which RNA sequencing data were available. The count fraction of the alternative alleles among the alternative and reference alleles together (φ) was determined for heterozygous individuals. A meta-analysis was performed to generate a meta-φ and P value for each SNP with a false discovery rate (FDR) correction for multiple comparisons. To further validate AI events, we explored them as a function of multiple additional OA features. RESULTS: We observed a total of 2,070 SNPs that consistently marked AI of 1,031 unique genes in articular cartilage. Of these genes, 49 were found to be significantly differentially expressed (fold change <0.5 or >2, FDR <0.05) between preserved and paired lesioned cartilage, and 18 had previously been reported to confer susceptibility to OA and/or related phenotypes. Moreover, we identified notable highly significant AI SNPs in the CRLF1, WWP2, and RPS3 genes that were related to multiple OA features. CONCLUSION: We present a framework and resulting data set for researchers in the OA research field to probe for disease-relevant genetic variation that affects gene expression in pivotal disease-affected tissue. This likely includes putative novel compelling OA risk genes such as CRLF1, WWP2, and RPS3.

Investigation of allele-specific expression of genes involved in adipogenesis and lipid metabolism suggests complex regulatory mechanisms of PPARGC1A expression in porcine fat tissues.

BACKGROUND: The expression of genes involved in regulating adipogenesis and lipid metabolism may affect economically important fatness traits in pigs. Allele-specific expression (ASE) reflects imbalance between allelic transcript levels and can be used to identify underlying cis-regulatory elements. ASE has not yet been intensively studied in pigs. The aim of this investigation was to analyze the differential allelic expression of four genes, PPARA, PPARG, SREBF1, and PPARGC1A, which are involved in the regulation of fat deposition in porcine subcutaneous and visceral fat and longissimus dorsi muscle. RESULTS: Quantification of allelic proportions by pyrosequencing revealed that both alleles of PPARG and SREBF1 are expressed at similar levels. PPARGC1A showed the greatest ASE imbalance in fat deposits in Polish Large White (PLW), Polish Landrace and Pietrain pigs; and PPARA in PLW pigs. Significant deviations of mean PPARGC1A allelic transcript ratio between cDNA and genomic DNA were detected in all tissues, with the most pronounced difference (p < 0.001) in visceral fat of PLW pigs. To search for potential cis-regulatory elements affecting ASE in the PPARGC1A gene we analyzed the effects of four SNPs (rs337351686, rs340650517, rs336405906 and rs345224049) in the promoter region, but none were associated with ASE in the breeds studied. DNA methylation analysis revealed significant CpG methylation differences between samples showing balanced (allelic transcript ratio ≈1) and imbalanced allelic expression for CpG site at the genomic position in chromosome 8 (SSC8): 18527678 in visceral fat (p = 0.017) and two CpG sites (SSC8:18525215, p = 0.030; SSC8:18525237, p = 0.031) in subcutaneous fat. CONCLUSIONS: Our analysis of differential allelic expression suggests that PPARGC1A is subjected to cis-regulation in porcine fat tissues. Further studies are necessary to identify other regulatory elements localized outside the PPARGC1A proximal promoter region.

Using Droplet Digital PCR to Analyze Allele-Specific RNA Expression.

Genome-wide association studies have discovered thousands of common alleles that associate with human phenotypes and disease. Many of these variants are in non-protein-coding (regulatory) regions and are believed to affect phenotypes by modifying gene expression. In any organism with a diploid genome, such as humans, measuring the expression of each allele of a gene provides a well-controlled way to identify allelic influences on that gene's expression. Here, we describe a protocol for precisely measuring the allele-specific expression of individual genes. This method targets the nucleotide differences between the two alleles of a gene within an individual and measures the "allelic skew," the extent to which one allele is expressed more than the other. We cover the design of effective assays, the optimization of reactions, and the interpretation of the resulting data.

Allelic Expression Imbalance Promoting a Mutant PEX6 Allele Causes Zellweger Spectrum Disorder.

Zellweger spectrum disorders (ZSDs) are autosomal-recessive disorders that are caused by defects in peroxisome biogenesis due to bi-allelic mutations in any of 13 different PEX genes. Here, we identified seven unrelated individuals affected with an apparent dominant ZSD in whom a heterozygous mutant PEX6 allele (c.2578C>T [p.Arg860Trp]) was overrepresented due to allelic expression imbalance (AEI). We demonstrated that AEI of PEX6 is a common phenomenon and is correlated with heterozygosity for a frequent variant in the 3' untranslated region (UTR) of the mutant allele, which disrupts the most distal of two polyadenylation sites. Asymptomatic parents, who were heterozygous for PEX c.2578C>T, did not show AEI and were homozygous for the 3' UTR variant. Overexpression models confirmed that the overrepresentation of the pathogenic PEX6 c.2578T variant compared to wild-type PEX6 c.2578C results in a peroxisome biogenesis defect and thus constitutes the cause of disease in the affected individuals. AEI promoting the overrepresentation of a mutant allele might also play a role in other autosomal-recessive disorders, in which only one heterozygous pathogenic variant is identified.

Allelic imbalance of somatic mutations in cancer genomes and transcriptomes.

Somatic mutations in cancer genomes often show allelic imbalance (AI) of mutation abundance between the genome and transcriptome, but there is not yet a systematic understanding of AI. In this study, we performed large-scale DNA and RNA AI analyses of >100,000 somatic mutations in >2,000 cancer specimens across five tumor types using the exome and transcriptome sequencing data of the Cancer Genome Atlas consortium. First, AI analysis of nonsense mutations and frameshift indels revealed that nonsense-mediated decay is typical in cancer genomes, and we identified the relationship between the extent of AI and the location of mutations in addition to the well-recognized 50-nt rules. Second, the AI with splice site mutations may reflect the extent of intron retention and is frequently observed in known tumor suppressor genes. For missense mutations, we observed that mutations frequently subject to AI are enriched to genes related to cancer, especially those of apoptosis and the extracellular matrix, and C:G > A:T transversions. Our results suggest that mutations in known cancer-related genes and their transcripts are subjected to different levels of transcriptional or posttranscriptional regulation compared to wildtype alleles and may add an additional regulatory layer to the functions of cancer-relevant genes.