Iris

We provide a statistical significance assessment for the presence of a cell type in the mixture by learning scores distributions for cell types in random mixtures. For each cell type X, we generate a random matrix as follows: In each reference data set we find all cell types corresponding to samples, except X and its parent or descendants (if X is CD8+ Tem cells, then we also exclude CD8+ T cells; if X is CD8+ T cells, we exclude all CD8+ cell types). We then use the same procedure we used for generating training samples, but adding an additional 5% random noise. The main difference here is that we randomly mix in all cell types (except X) and not just a small subset. We then run the xCell pipeline for these random mixtures. In most cell types the produced scores show similarity to a beta distribution; thus, using the fitdistr function from the MASS package, we fit such a distribution for each of the mixtures we generated (e.g., for a mixture excluding cell type X we fit a beta distribution for cell type X). In five of the cell types the scores from the random mixtures consistently produced 0; thus, we define those distributions as constant 0.001 (Additional file 2: Figure S7). Given an input data set, we can now calculate a p value for each xCell score with the null hypothesis that the cell type is not present in the mixture. The actual distributions we use to calculate the p values are combinations of those learned from FANTOM5, Blueprint, and ENCODE for sequencing-based input, and IRIS, HPCA, and Novershtern for microarray-based input. The p value for a score of a cell type in a sample is the chance of the region in the distribution of the corresponding cell type to exceed the score. In the testing samples we used a threshold of 20% to define a non-significant score. We used this threshold to have a trade-off between detecting the non-negligible scores of cell types not in the mixture and not detecting scores of cell type in the mixture, thus affecting the power of estimating the underlying cell type fractions (Additional file 4).

To facilitate a comfortable start for users, we have included several example data sets that can be found in the compressed file (folder: examples). In light of the versatile application of Instant Clue, we have included fully documented step-by-step data analysis procedure in the tutorial (http://www.instantclue.uni-koeln.de/tutorials.html) of various different type of data sets: (i) Body weight measurements of people of different health condition and age. (ii) mass spectrometry based immunoprecipitation data published recently to identify interaction partners of a protease dead mutant of Presenilins-associated rhomboid-like protein (PARL)^{18 (link),19 (link)}. (iii) optical recording of Pro-opiomelanocortin (POMC) neuron activity (time series data)^{20 (link)}. (iv) iris data set^{21 (link)}. (v) wine quality data set for supervised learning^{22 (link)}. As the tutorial will be extended continuously, we will also add more example data.

Lists of photography instances were generated, and these lists were used to investigate all photography orders recorded in the Epic EMR utilized by WVU Hospitals between January 2017 and June 2019. Photography orders that were unfulfilled (due to premature order placement by clinicians, for instance) were excluded from the study. Patient information was de-identified, and spreadsheets in Microsoft Excel were created to collect and organize the data.
Each valid photography order was investigated in the following fashion. First, the IRIS results adjoined to patients’ charts for the photography order in question would be accessed. The gradeability and presence of pathology would be recorded (specifically noting DR as mild, moderate, severe, or proliferative and DME as mild, moderate, or severe). If the screening results indicated suspicion for pathology, further investigation was conducted. Date of birth, the time that had passed since their diabetes diagnosis, their diabetes classification (Type 1 or Type 2), and HbA1c within 3-months of their photography date were all collected. Patient receipt of their results (either through record of PCP communications or indications that patients had read their results via the patient-accessible WVU MyChart system) was recorded, and whether or not an appointment was subsequently set and maintained (within 12 months of the photography order date or prior to a future repeat screening with their PCP) was also noted. Using patients’ home addresses, distances from the WVU Eye Institute to patients’ hometowns were recorded using Google Maps driving estimates. The results of patients’ dilated eye exams were recorded (noting severity as mild, moderate, severe, or proliferative for DR and absent or present for DME). Where feasible, these data were acquired from offices outside of WVU Medicine by either viewing documentation that had already been uploaded to the Epic EMR by patients or their providers or by contacting these offices directly where references in PCP notes indicated completion of ophthalmic follow-up outside WVU Medicine and permission had been granted.

This study enrolled consecutive patients aged >50 years who visited the ophthalmology department of Ramathibodi Hospital, Mahidol University, Thailand, with a complaint of symptomatic cataract with and without the signs of pPEX between April 2018 and December 2018.
pPEX was defined as the presence of the following signs without clinically identifiable PXM on the anterior lens capsule or pupillary margin in either eye: (1) pigmented spoke-wheel deposition (P) on the anterior lens capsule (Fig 1A), (2) faint central disc (D) present within the photopic pupil (Fig 1B), (3) midperiphery cleft/lacunae (C; Fig 1C), and (4) white-spoke pattern (W) noted at the midperiphery of the lens capsule (Fig 1D). We hypothesized that P, D, and C originated from continuous rubbing of PXM on the posterior iris against anterior lens, whereas W represents early stage of PXM formation.
We included patients who underwent complete ocular examination. This study excluded patients with a history of eye trauma, uveitis, pigmentary dispersion, media opacities, conditions that affected the anterior chamber and angle examination, and laser or surgical treatment; those with the presence of PXM on the pupillary margin or anterior lens capsule; and those who were unwilling to participate.
The enrolled patients were examined using slit-lamp biomicroscopy as part of a detailed preoperative examination. The anterior capsular surface was photographed using a photo slit lamp (Haag-Streit BX900, Haag-Streit AG, Switzerland). Clinical data, including age, sex, associated eye disease, and true exfoliation syndrome (TEX) stage (if present), were collected.

Crystalline lens samples were obtained from conventional continuous curvilinear capsulorhexis in ARC patients (n = 10, each sample with three replicates), with no vascular contact or damage to the iris or other intraocular tissues. Patients with complex cataracts due to high myopia, trauma, uveitis, glaucoma, or other systemic diseases, such as hypertension and diabetes, were excluded from the study. Transparent lenses removed from normal subjects with shallow anterior chambers served as controls (n = 2, each sample with three replicates). The “shallow anterior chamber” is only a potential risk factor for glaucoma, but there is currently no related disease, and it also belongs to normal people. In this group, the clinical option was to deepen the anterior chamber through preventive lens surgery to remove the potential glaucoma risk. Signed informed consent was obtained from all patients. Detailed clinical data for each individual human subject are shown in Table S1. This study was approved by the Ethics Committee of Shanghai East Hospital, School of Medicine, Tongji University ([2021] Audit Research No. 79).