Comparative evaluation of the Minimally-Invasive Karyotyping (MINK) algorithm for non-invasive prenatal testing.

Minimally Invasive Karyotyping (MINK) was communicated in 2009 as a novel method for the non-invasive detection of fetal copy number anomalies in maternal plasma DNA. The original manuscript illustrated the potential of MINK using a model system in which fragmented genomic DNA obtained from a trisomy 21 male individual was mixed with that of his karyotypically normal mother at dilutions representing fetal fractions found in maternal plasma. Although it has been previously shown that MINK is able to non-invasively detect fetal microdeletions, its utility for aneuploidy detection in maternal plasma has not previously been demonstrated. The current study illustrates the ability of MINK to detect common aneuploidy in early gestation, compares its performance to other published third party methods (and related software packages) for prenatal aneuploidy detection and evaluates the performance of these methods across a range of sequencing read inputs. Plasma samples were obtained from 416 pregnant women between gestational weeks 8.1 and 34.4. Shotgun DNA sequencing was performed and data analyzed using MINK RAPIDR and WISECONDOR. MINK performed with greater accuracy than RAPIDR and WISECONDOR, correctly identifying 60 out of 61 true trisomy cases, and reporting only one false positive in 355 normal pregnancies. Significantly, MINK achieved accurate detection of trisomy 21 using just 2 million aligned input reads, whereas WISECONDOR required 6 million reads and RAPIDR did not achieve complete accuracy at any read input tested. In conclusion, we demonstrate that MINK provides an analysis pipeline for the detection of fetal aneuploidy in samples of maternal plasma DNA.

Correction: High Levels of Sample-to-Sample Variation Confound Data Analysis for Non-Invasive Prenatal Screening of Fetal Microdeletions.

[This corrects the article DOI: 10.1371/journal.pone.0153182.].

High Levels of Sample-to-Sample Variation Confound Data Analysis for Non-Invasive Prenatal Screening of Fetal Microdeletions.

Our goal was to test the hypothesis that inter-individual genomic copy number variation in control samples is a confounding factor in the non-invasive prenatal detection of fetal microdeletions via the sequence-based analysis of maternal plasma DNA. The database of genomic variants (DGV) was used to determine the "Genomic Variants Frequency" (GVF) for each 50kb region in the human genome. Whole genome sequencing of fifteen karyotypically normal maternal plasma and six CVS DNA controls samples was performed. The coefficient of variation of relative read counts (cv.RTC) for these samples was determined for each 50kb region. Maternal plasma from two pregnancies affected with a chromosome 5p microdeletion was also sequenced, and analyzed using the GCREM algorithm. We found strong correlation between high variance in read counts and GVF amongst controls. Consequently we were unable to confirm the presence of the microdeletion via sequencing of maternal plasma samples obtained from two sequential affected pregnancies. Caution should be exercised when performing NIPT for microdeletions. It is vital to develop our understanding of the factors that impact the sensitivity and specificity of these approaches. In particular, benign copy number variation amongst controls is a major confounder, and their effects should be corrected bioinformatically.

High resolution non-invasive detection of a fetal microdeletion using the GCREM algorithm.

BACKGROUND/OBJECTIVE: The non-invasive prenatal detection of fetal microdeletions becomes increasingly challenging as the size of the mutation decreases, with current practical lower limits in the range of a few megabases. Our goals were to explore the lower limits of microdeletion size detection via non-invasive prenatal tests using Minimally Invasive Karyotyping (MINK) and introduce/evaluate a novel statistical approach we recently developed called the GC Content Random Effect Model (GCREM). METHODS: Maternal plasma was obtained from a pregnancy affected by a 4.2-Mb fetal microdeletion and three normal controls. Plasma DNA was subjected to capture an 8-Mb sequence spanning the breakpoint region and sequence. Data were analyzed with our published method, MINK, and a new method called GCREM. RESULTS: The 8-Mb capture segment was divided into either 38 or 76 non-overlapping regions of 200 and 100 Kb, respectively. At 200 Kb resolution, using GCREM (but not MINK), we obtained significant adjusted p-values for all 20 regions overlapping the deleted sequence, and non-significant p-values for all 18 reference regions. At 100 Kb resolution, GCREM identified significant adjusted p-values for all but one 100-Kb region located inside the deleted region. CONCLUSION: Targeted sequencing and GCREM analysis may enable cost effective detection of fetal microdeletions and microduplications at high resolution.

EnzyMiner: automatic identification of protein level mutations and their impact on target enzymes from PubMed abstracts.

BACKGROUND: A better understanding of the mechanisms of an enzyme's functionality and stability, as well as knowledge and impact of mutations is crucial for researchers working with enzymes. Though, several of the enzymes' databases are currently available, scientific literature still remains at large for up-to-date source of learning the effects of a mutation on an enzyme. However, going through vast amounts of scientific documents to extract the information on desired mutation has always been a time consuming process. In this paper, therefore, we describe an unique method, termed as EnzyMiner, which automatically identifies the PubMed abstracts that contain information on the impact of a protein level mutation on the stability and/or the activity of a given enzyme. RESULTS: We present an automated system which identifies the abstracts that contain an amino-acid-level mutation and then classifies them according to the mutation's effect on the enzyme. In the case of mutation identification, MuGeX, an automated mutation-gene extraction system has an accuracy of 93.1% with a 91.5 F-measure. For impact analysis, document classification is performed to identify the abstracts that contain a change in enzyme's stability or activity resulting from the mutation. The system was trained on lipases and tested on amylases with an accuracy of 85%. CONCLUSION: EnzyMiner identifies the abstracts that contain a protein mutation for a given enzyme and checks whether the abstract is related to a disease with the help of information extraction and machine learning techniques. For disease related abstracts, the mutation list and direct links to the abstracts are retrieved from the system and displayed on the Web. For those abstracts that are related to non-diseases, in addition to having the mutation list, the abstracts are also categorized into two groups. These two groups determine whether the mutation has an effect on the enzyme's stability or functionality followed by displaying these on the web.