Jaime I. Davila
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Jaime I. Davila has a total of 174 co-authors. Here are the top co-authors by number of shared publications:
- Chen Wang
Mayo Clinic
6 shared publications - Kevin C. Halling
Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN
6 shared publications Numrah Fadra
Department of Quantitative Health Sciences Research, Mayo Clinic, Rochester, MN
4 shared publicationsBeth A. Pitel
Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN
3 shared publications- Jin Jen
Mayo Clinic
3 shared publications - Mazen A. Atiq
Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN
3 shared publications
Latest Publications
- Open Access
- 3 June 2026
- Published by Elsevier in Translational Oncology
- Vol. 70, 102812
RNA-Based and DNA-Based Next-Generation Sequencing of KIT and PDGFRA Mutations in Gastrointestinal Stromal Tumors: Analytical Performance and Epidemiologic Insights
- 3 June 2026
- Published by Elsevier in Translational Oncology
- Vol. 70, 102812
Open AccessRNA-Based and DNA-Based Next-Generation Sequencing of KIT and PDGFRA Mutations in Gastrointestinal Stromal Tumors: Analytical Performance and Epidemiologic Insights
- Open Access
- 19 March 2026
- Published by Elsevier in SSRN Electronic Journal
RNA-Based and DNA-Based Next-Generation Sequencing of KIT and PDGFRA Mutations in Gastrointestinal Stromal Tumors: Analytical Performance and Epidemiologic Insights
Mutations in the KIT and PDGFRA proto-oncogenes are key drivers in gastrointestinal stromal tumors (GIST) and guide targeted therapy. While DNA-based next-generation sequencing (NGS) is standard for mutation detection, RNA sequencing (RNA-Seq) may offer complementary advantages. This study evaluates RNA-Seq performance and presents molecular epidemiologic data from a large DNA-sequenced GIST cohort. RNA-Seq was performed on 24 GIST cases previously analyzed by DNA-based NGS (16 KIT mutants, 5 PDGFRA mutants, 3 wild-type) using the Agilent SureSelectXT RNA Direct Library Prep Kit and an in-house analysis pipeline. RNA-Seq was successful in 21 cases, identifying 13 of 14 KIT mutations and all PDGFRA mutations, including single nucleotide variants, insertions, and in-frame deletions (3–27 bp). One complex 45 bp deletion-insertion was not detected, though low-level evidence (<10% of reads) was present. Separately, DNA-based NGS results from 579 GIST cases were reviewed to assess mutation prevalence and clinicopathologic associations. Mutations were found in 83.2% of cases (403 KIT, 79 PDGFRA), with secondary KIT mutations in 6.0%. Mutation site correlated with tumor location and patient age; secondary mutations were more frequent in non-gastric tumors. RNA-Seq demonstrates high accuracy for detecting clinically relevant KIT and PDGFRA mutations and may complement DNA-based profiling. The DNA cohort provides broader context for mutation prevalence and patterns in clinical practice.
- 19 March 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessRNA-Based and DNA-Based Next-Generation Sequencing of KIT and PDGFRA Mutations in Gastrointestinal Stromal Tumors: Analytical Performance and Epidemiologic Insights
Mutations in the KIT and PDGFRA proto-oncogenes are key drivers in gastrointestinal stromal tumors (GIST) and guide targeted therapy. While DNA-based next-generation sequencing (NGS) is standard for mutation detection, RNA sequencing (RNA-Seq) may offer complementary advantages. This study evaluates RNA-Seq performance and presents molecular epidemiologic data from a large DNA-sequenced GIST cohort. RNA-Seq was performed on 24 GIST cases previously analyzed by DNA-based NGS (16 KIT mutants, 5 PDGFRA mutants, 3 wild-type) using the Agilent SureSelectXT RNA Direct Library Prep Kit and an in-house analysis pipeline. RNA-Seq was successful in 21 cases, identifying 13 of 14 KIT mutations and all PDGFRA mutations, including single nucleotide variants, insertions, and in-frame deletions (3–27 bp). One complex 45 bp deletion-insertion was not detected, though low-level evidence (<10% of reads) was present. Separately, DNA-based NGS results from 579 GIST cases were reviewed to assess mutation prevalence and clinicopathologic associations. Mutations were found in 83.2% of cases (403 KIT, 79 PDGFRA), with secondary KIT mutations in 6.0%. Mutation site correlated with tumor location and patient age; secondary mutations were more frequent in non-gastric tumors. RNA-Seq demonstrates high accuracy for detecting clinically relevant KIT and PDGFRA mutations and may complement DNA-based profiling. The DNA cohort provides broader context for mutation prevalence and patterns in clinical practice.
- 21 February 2025
- Published by Springer Nature
- p. 65-74
Cancer and Tissue Prediction Using Mutational Signatures in Highly Mutated Cancers
- 21 February 2025
- Published by Springer Nature
- p. 65-74
Cancer and Tissue Prediction Using Mutational Signatures in Highly Mutated Cancers
- 7 November 2024
- Published by Elsevier in The Journal of Molecular Diagnostics
- Vol. 27 (1) , 74-95
SARCP, a Clinical Next-Generation Sequencing Assay for the Detection of Gene Fusions in Sarcomas
- 7 November 2024
- Published by Elsevier in The Journal of Molecular Diagnostics
- Vol. 27 (1) , 74-95
SARCP, a Clinical Next-Generation Sequencing Assay for the Detection of Gene Fusions in Sarcomas
- 1 April 2023
- Published by SAGE Publications in Journal of Computational Biology
- Vol. 30 (4) , 366-375
Excerno: Using Mutational Signatures in Sequencing Data to Filter False Variants Caused by Clinical Archival
The accurate detection of point mutations from pathology slides using sequencing data is of great importance in cancer genomics and precision oncology. Formalin-fixation paraffin-embedding (FFPE) is a widely used technique to preserve pathology tissues. The FFPE process introduces artificial C > T mutations in next-generation sequencing, so we set out to develop excerno , a method to score and filter such spurious variants. FFPE mutational artifacts follow a mutational signature. By using the FFPE signature and Bayes' formula, we can calculate the probability of a mutation resulting from the FFPE process and use this probability to filter FFPE variants. We implement this method as the excerno R package. We tested excerno by simulating mutations across all 60-baseline mutational signatures from the Catalog of Somatic Mutations in Cancer (COSMIC) and combining them with mutations following the FFPE mutational signature. The sensitivity and specificity of excerno are adversely affected by the cosine similarity between the baseline and FFPE signatures ( c o s F F P E ). Higher percentages of FFPE mutations ( p c t F F P E ) result in increased sensitivity and reduced specificity. The specificity and sensitivity of excerno can be predicted as linear model with an interaction term using c o s F F P E and p c t F F P E , with an R 2 o f 0 . 8 4 and 0.79, respectively. Finally, we tested excerno using six RNA sequencing cancer samples and observed concordant trends of specificity and sensitivity with respect to our simulated data. The excerno R package can be used to annotate and filter FFPE-induced mutations in cancer genomics. Our method is adversely affected by c o s F F P E and p c t F F P E .
- 1 April 2023
- Published by SAGE Publications in Journal of Computational Biology
- Vol. 30 (4) , 366-375
Excerno: Using Mutational Signatures in Sequencing Data to Filter False Variants Caused by Clinical Archival
The accurate detection of point mutations from pathology slides using sequencing data is of great importance in cancer genomics and precision oncology. Formalin-fixation paraffin-embedding (FFPE) is a widely used technique to preserve pathology tissues. The FFPE process introduces artificial C > T mutations in next-generation sequencing, so we set out to develop excerno , a method to score and filter such spurious variants. FFPE mutational artifacts follow a mutational signature. By using the FFPE signature and Bayes' formula, we can calculate the probability of a mutation resulting from the FFPE process and use this probability to filter FFPE variants. We implement this method as the excerno R package. We tested excerno by simulating mutations across all 60-baseline mutational signatures from the Catalog of Somatic Mutations in Cancer (COSMIC) and combining them with mutations following the FFPE mutational signature. The sensitivity and specificity of excerno are adversely affected by the cosine similarity between the baseline and FFPE signatures ( c o s F F P E ). Higher percentages of FFPE mutations ( p c t F F P E ) result in increased sensitivity and reduced specificity. The specificity and sensitivity of excerno can be predicted as linear model with an interaction term using c o s F F P E and p c t F F P E , with an R 2 o f 0 . 8 4 and 0.79, respectively. Finally, we tested excerno using six RNA sequencing cancer samples and observed concordant trends of specificity and sensitivity with respect to our simulated data. The excerno R package can be used to annotate and filter FFPE-induced mutations in cancer genomics. Our method is adversely affected by c o s F F P E and p c t F F P E .
- 19 October 2022
- Published by Springer Nature
- p. 29-37
Excerno: Filtering Mutations Caused by the Clinical Archival Process in Sequencing Data
- 19 October 2022
- Published by Springer Nature
- p. 29-37
Excerno: Filtering Mutations Caused by the Clinical Archival Process in Sequencing Data
- Open Access
- 30 July 2022
- Published by Springer Nature in BMC Medical Genomics
- Vol. 15 (1) , 1-12
PO2RDF: representation of real-world data for precision oncology using resource description framework
Background: Next-generation sequencing provides comprehensive information about individuals’ genetic makeup and is commonplace in precision oncology practice. Due to the heterogeneity of individual patient’s disease conditions and treatment journeys, not all targeted therapies were initiated despite actionable mutations. To better understand and support the clinical decision-making process in precision oncology, there is a need to examine real-world associations between patients’ genetic information and treatment choices.Methods: To fill the gap of insufficient use of real-world data (RWD) in electronic health records (EHRs), we generated a single Resource Description Framework (RDF) resource, called PO2RDF (precision oncology to RDF), by integrating information regarding genes, variants, diseases, and drugs from genetic reports and EHRs.Results: There are a total 2,309,014 triples contained in the PO2RDF. Among them, 32,815 triples are related to Gene, 34,695 triples are related to Variant, 8,787 triples are related to Disease, 26,154 triples are related to Drug. We performed two use case analyses to demonstrate the usability of the PO2RDF: (1) we examined real-world associations between EGFR mutations and targeted therapies to confirm existing knowledge and detect off-label use. (2) We examined differences in prognosis for lung cancer patients with/without TP53 mutations.Conclusions: In conclusion, our work proposed to use RDF to organize and distribute clinical RWD that is otherwise inaccessible externally. Our work serves as a pilot study that will lead to new clinical applications and could ultimately stimulate progress in the field of precision oncology.
- 30 July 2022
- Published by Springer Nature in BMC Medical Genomics
- Vol. 15 (1) , 1-12
Open AccessPO2RDF: representation of real-world data for precision oncology using resource description framework
Background: Next-generation sequencing provides comprehensive information about individuals’ genetic makeup and is commonplace in precision oncology practice. Due to the heterogeneity of individual patient’s disease conditions and treatment journeys, not all targeted therapies were initiated despite actionable mutations. To better understand and support the clinical decision-making process in precision oncology, there is a need to examine real-world associations between patients’ genetic information and treatment choices.Methods: To fill the gap of insufficient use of real-world data (RWD) in electronic health records (EHRs), we generated a single Resource Description Framework (RDF) resource, called PO2RDF (precision oncology to RDF), by integrating information regarding genes, variants, diseases, and drugs from genetic reports and EHRs.Results: There are a total 2,309,014 triples contained in the PO2RDF. Among them, 32,815 triples are related to Gene, 34,695 triples are related to Variant, 8,787 triples are related to Disease, 26,154 triples are related to Drug. We performed two use case analyses to demonstrate the usability of the PO2RDF: (1) we examined real-world associations between EGFR mutations and targeted therapies to confirm existing knowledge and detect off-label use. (2) We examined differences in prognosis for lung cancer patients with/without TP53 mutations.Conclusions: In conclusion, our work proposed to use RDF to organize and distribute clinical RWD that is otherwise inaccessible externally. Our work serves as a pilot study that will lead to new clinical applications and could ultimately stimulate progress in the field of precision oncology.
- 23 March 2022
- Published by Elsevier in Cancer Genetics
- Vol. 260-261, 12
35. Atypical FISH patterns clarified by RNAseq in solid tumor specimens
- 23 March 2022
- Published by Elsevier in Cancer Genetics
- Vol. 260-261, 12
35. Atypical FISH patterns clarified by RNAseq in solid tumor specimens
- 6 August 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ACM Transactions on Computational Biology and Bioinformatics
- Vol. 18 (4) , 1620-1631
Semi-Supervised Topological Analysis for Elucidating Hidden Structures in High-Dimensional Transcriptome Datasets
Topological data analysis (TDA) is a powerful method for reducing data dimensionality, mining underlying data relationships, and intuitively representing the data structure. The Mapper algorithm is one such tool that projects high-dimensional data to 1-dimensional space by using a filter function that is subsequently used to reconstruct the data topology relationships. However, domain context information and prior knowledge have not been considered in current TDA modeling frameworks. Here, we report the development and evaluation of a semisupervised topological analysis (STA) framework that incorporates discrete or continuously labeled data points and selects the most relevant filter functions accordingly. We validate the proposed STA framework with simulation data and then apply it to samples from Genotype-Tissue Expression data and ovarian cancer transcriptome datasets. The graphs generated by STA for these 2 datasets, based on gene expression profiles, are consistent with prior knowledge, thereby supporting the effectiveness of the proposed framework.
- 6 August 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ACM Transactions on Computational Biology and Bioinformatics
- Vol. 18 (4) , 1620-1631
Semi-Supervised Topological Analysis for Elucidating Hidden Structures in High-Dimensional Transcriptome Datasets
Topological data analysis (TDA) is a powerful method for reducing data dimensionality, mining underlying data relationships, and intuitively representing the data structure. The Mapper algorithm is one such tool that projects high-dimensional data to 1-dimensional space by using a filter function that is subsequently used to reconstruct the data topology relationships. However, domain context information and prior knowledge have not been considered in current TDA modeling frameworks. Here, we report the development and evaluation of a semisupervised topological analysis (STA) framework that incorporates discrete or continuously labeled data points and selects the most relevant filter functions accordingly. We validate the proposed STA framework with simulation data and then apply it to samples from Genotype-Tissue Expression data and ovarian cancer transcriptome datasets. The graphs generated by STA for these 2 datasets, based on gene expression profiles, are consistent with prior knowledge, thereby supporting the effectiveness of the proposed framework.
- Open Access
- 22 June 2021
- Published by Springer Nature in BMC Medical Genomics
- Vol. 14 (1) , 1-6
Frequent POLE-driven hypermutation in ovarian endometrioid cancer revealed by mutational signatures in RNA sequencing
Background: DNA polymerase epsilon (POLE) is encoded by the POLE gene, and POLE-driven tumors are characterized by high mutational rates. POLE-driven tumors are relatively common in endometrial and colorectal cancer, and their presence is increasingly recognized in ovarian cancer (OC) of endometrioid type. POLE-driven cases possess an abundance of TCT > TAT and TCG > TTG somatic mutations characterized by mutational signature 10 from the Catalog of Somatic Mutations in Cancer (COSMIC). By quantifying the contribution of COSMIC mutational signature 10 in RNA sequencing (RNA-seq) we set out to identify POLE-driven tumors in a set of unselected Mayo Clinic OC. Methods: Mutational profiles were calculated using expressed single-nucleotide variants (eSNV) in the Mayo Clinic OC tumors (n = 195), The Cancer Genome Atlas (TCGA) OC tumors (n = 419), and the Genotype-Tissue Expression (GTEx) normal ovarian tissues (n = 84). Non-negative Matrix Factorization (NMF) of the mutational profiles inferred the contribution per sample of four distinct mutational signatures, one of which corresponds to COSMIC mutational signature 10. Results: In the Mayo Clinic OC cohort we identified six tumors with a predicted contribution from COSMIC mutational signature 10 of over five mutations per megabase. These six cases harbored known POLE hotspot mutations (P286R, S297F, V411L, and A456P) and were of endometrioid histotype (P = 5e−04). These six tumors had an early onset (average age of patients at onset, 48.33 years) when compared to non-POLE endometrioid OC cohort (average age at onset, 60.13 years; P = .008). Samples from TCGA and GTEx had a low COSMIC signature 10 contribution (median 0.16 mutations per megabase; maximum 1.78 mutations per megabase) and carried no POLE hotspot mutations. Conclusions: From the largest cohort of RNA-seq from endometrioid OC to date (n = 53), we identified six hypermutated samples likely driven by POLE (frequency, 11%). Our result suggests the clinical need to screen for POLE driver mutations in endometrioid OC, which can guide enrollment in immunotherapy clinical trials.
- 22 June 2021
- Published by Springer Nature in BMC Medical Genomics
- Vol. 14 (1) , 1-6
Open AccessFrequent POLE-driven hypermutation in ovarian endometrioid cancer revealed by mutational signatures in RNA sequencing
Background: DNA polymerase epsilon (POLE) is encoded by the POLE gene, and POLE-driven tumors are characterized by high mutational rates. POLE-driven tumors are relatively common in endometrial and colorectal cancer, and their presence is increasingly recognized in ovarian cancer (OC) of endometrioid type. POLE-driven cases possess an abundance of TCT > TAT and TCG > TTG somatic mutations characterized by mutational signature 10 from the Catalog of Somatic Mutations in Cancer (COSMIC). By quantifying the contribution of COSMIC mutational signature 10 in RNA sequencing (RNA-seq) we set out to identify POLE-driven tumors in a set of unselected Mayo Clinic OC. Methods: Mutational profiles were calculated using expressed single-nucleotide variants (eSNV) in the Mayo Clinic OC tumors (n = 195), The Cancer Genome Atlas (TCGA) OC tumors (n = 419), and the Genotype-Tissue Expression (GTEx) normal ovarian tissues (n = 84). Non-negative Matrix Factorization (NMF) of the mutational profiles inferred the contribution per sample of four distinct mutational signatures, one of which corresponds to COSMIC mutational signature 10. Results: In the Mayo Clinic OC cohort we identified six tumors with a predicted contribution from COSMIC mutational signature 10 of over five mutations per megabase. These six cases harbored known POLE hotspot mutations (P286R, S297F, V411L, and A456P) and were of endometrioid histotype (P = 5e−04). These six tumors had an early onset (average age of patients at onset, 48.33 years) when compared to non-POLE endometrioid OC cohort (average age at onset, 60.13 years; P = .008). Samples from TCGA and GTEx had a low COSMIC signature 10 contribution (median 0.16 mutations per megabase; maximum 1.78 mutations per megabase) and carried no POLE hotspot mutations. Conclusions: From the largest cohort of RNA-seq from endometrioid OC to date (n = 53), we identified six hypermutated samples likely driven by POLE (frequency, 11%). Our result suggests the clinical need to screen for POLE driver mutations in endometrioid OC, which can guide enrollment in immunotherapy clinical trials.