Active deep learning to detect demographic traits in free-form clinical notes.

The free-form portions of clinical notes are a significant source of information for research, but before they can be used, they must be de-identified to protect patients' privacy. De-identification efforts have focused on known identifier types (names, ages, dates, addresses, ID's, etc.). However, a note can contain residual "Demographic Traits" (DTs), unique enough to re-identify the patient when combined with other such facts. Here we examine whether any residual risks remain after removing these identifiers. After manually annotating over 140,000 words worth of medical notes, we found no remaining directly identifying information, and a low prevalence of demographic traits, such as marital status or housing type. We developed an annotation guide to the discovered Demographic Traits (DTs) and used it to label MIMIC-III and i2b2-2006 clinical notes as test sets. We then designed a "bootstrapped" active learning iterative process for identifying DTs: we tentatively labeled as positive all sentences in the DT-rich note sections, used these to train a binary classifier, manually corrected acute errors, and retrained the classifier. This train-and-correct process may be iterated. Our active learning process significantly improved the classifier's accuracy. Moreover, our BERT-based model outperformed non-neural models when trained on both tentatively labeled data and manually relabeled examples. To facilitate future research and benchmarking, we also produced and made publicly available our human annotated DT-tagged datasets. We conclude that directly identifying information is virtually non-existent in the multiple medical note types we investigated. Demographic traits are present in medical notes, but can be detected with high accuracy using a cost-effective human-in-the-loop active learning process, and redacted if desired.2.

Customization scenarios for de-identification of clinical notes.

BACKGROUND: Automated machine-learning systems are able to de-identify electronic medical records, including free-text clinical notes. Use of such systems would greatly boost the amount of data available to researchers, yet their deployment has been limited due to uncertainty about their performance when applied to new datasets. OBJECTIVE: We present practical options for clinical note de-identification, assessing performance of machine learning systems ranging from off-the-shelf to fully customized. METHODS: We implement a state-of-the-art machine learning de-identification system, training and testing on pairs of datasets that match the deployment scenarios. We use clinical notes from two i2b2 competition corpora, the Physionet Gold Standard corpus, and parts of the MIMIC-III dataset. RESULTS: Fully customized systems remove 97-99% of personally identifying information. Performance of off-the-shelf systems varies by dataset, with performance mostly above 90%. Providing a small labeled dataset or large unlabeled dataset allows for fine-tuning that improves performance over off-the-shelf systems. CONCLUSION: Health organizations should be aware of the levels of customization available when selecting a de-identification deployment solution, in order to choose the one that best matches their resources and target performance level.

A 1.375-approximation algorithm for sorting by transpositions.

Sorting permutations by transpositions is an important problem in genome rearrangements. A transposition is a rearrangement operation in which a segment is cut out of the permutation and pasted in a different location. The complexity of this problem is still open and it has been a 10-year-old open problem to improve the best known 1.5-approximation algorithm. In this paper, we provide a 1.375-approximation algorithm for sorting by transpositions. The algorithm is based on a new upper bound on the diameter of 3-permutations. In addition, we present some new results regarding the transposition diameter: we improve the lower bound for the transposition diameter of the symmetric group and determine the exact transposition diameter of simple permutations.

Handling long targets and errors in sequencing by hybridization.

Sequencing by hybridization (SBH) is a DNA sequencing technique, in which the sequence is reconstructed using its k-mer content. This content, which is called the spectrum of the sequence, is obtained by hybridization to a universal DNA array. Standard universal arrays contain all k-mers for some fixed k, typically 8 to 10. Currently, in spite of its promise and elegance, SBH is not competitive with standard gel-based sequencing methods. This is due to two main reasons: lack of tools to handle realistic levels of hybridization errors and an inherent limitation on the length of uniquely reconstructible sequence by standard universal arrays. In this paper, we deal with both problems. We introduce a simple polynomial reconstruction algorithm which can be applied to spectra from standard arrays and has provable performance in the presence of both false negative and false positive errors. We also propose a novel design of chips containing universal bases that differs from the one proposed by Preparata et al. (1999). We give a simple algorithm that uses spectra from such chips to reconstruct with high probability random sequences of length lower only by a squared log factor compared to the information theoretic bound. Our algorithm is very robust to errors and has a provable performance even if there are both false negative and false positive errors. Simulations indicate that its sensitivity to errors is also very small in practice.

Towards optimally multiplexed applications of universal arrays.

We study a design and optimization problem that occurs, for example, when single nucleotide polymorphisms (SNPs) are to be genotyped using a universal DNA tag array. The problem of optimizing the universal array to avoid disruptive cross-hybridization between universal components of the system was addressed in previous work. Cross-hybridization can, however, also occur assay specifically, due to unwanted complementarity involving assay-specific components. Here we examine the problem of identifying the most economic experimental configuration of the assay-specific components that avoids cross-hybridization. Our formalization translates this problem into the problem of covering the vertices of one side of a bipartite graph by a minimum number of balanced subgraphs of maximum degree 1. We show that the general problem is NP-complete. However, in the real biological setting, the vertices that need to be covered have degrees bounded by d. We exploit this restriction and develop an O(d)-approximation algorithm for the problem. We also give an O(d)-approximation for a variant of the problem in which the covering subgraphs are required to be vertex disjoint. In addition, we propose a stochastic model for the input data and use it to prove a lower bound on the cover size. We complement our theoretical analysis by implementing two heuristic approaches and testing their performance on synthetic data as well as on simulated SNP data.

Haplotype inference constrained by plausible haplotype data.

The haplotype inference problem (HIP) asks to find a set of haplotypes which resolve a given set of genotypes. This problem is important in practical fields such as the investigation of diseases or other types of genetic mutations. In order to find the haplotypes which are as close as possible to the real set of haplotypes that comprise the genotypes, two models have been suggested which are by now well-studied: The perfect phylogeny model and the pure parsimony model. All known algorithms up till now for haplotype inference may find haplotypes that are not necessarily plausible, i.e., very rare haplotypes or haplotypes that were never observed in the population. In order to overcome this disadvantage, we study in this paper, a new constrained version of HIP under the above-mentioned models. In this new version, a pool of plausible haplotypes H is given together with the set of genotypes G, and the goal is to find a subset H ⊆ H that resolves G. For constrained perfect phlogeny haplotyping (CPPH), we provide initial insights and polynomial-time algorithms for some restricted cases of the problem. For constrained parsimony haplotyping (CPH), we show that the problem is fixed parameter tractable when parameterized by the size of the solution set of haplotypes.