Journal: Thoracic Cancer

Article Title: Effect of lymph node examined count on accurate staging and survival of resected esophageal cancer

doi: 10.1111/1759-7714.13056

Figure Lengend Snippet: Co‐plot of odds ratios (ORs, upper row) and Locally Weighted Scatterplot Smoothing curves of stage migration and determination of structural break points with the use of the Chow test (bottom row). The fitting bandwidth was 0.6. Each dot in the co‐plot represents an OR of a specific lymph node examined (LNE, vacant if case number < 10) from logistic regression analysis. ( a ) Overall patients, ( b ) adenocarcinoma patients, and ( c ) esophageal squamous cell carcinoma (ESCC) patients from the Surveillance, Epidemiology, and End Results (SEER) database; and ( d ) ESCC patients from the Thoracic Surgery Department of a single institution (SI). ( ) T1, ( ) T2, ( ) T3, and ( ) T4.

Article Snippet: **Curves were generated using odds ratios (ORs, each LNE count compared to 1 LNE as a reference) in logistic regression analysis and hazard ratios (HRs) in Cox regression analysis using Locally Weighted Scatterplot Smoothing (LOWESS, Stata 12.0) with a bandwidth of 0.6.

Techniques: Migration

Journal: Thoracic Cancer

Article Title: Effect of lymph node examined count on accurate staging and survival of resected esophageal cancer

doi: 10.1111/1759-7714.13056

Figure Lengend Snippet: Co‐plot of hazard ratios (HRs, upper row) and Locally Weighted Scatterplot Smoothing curves of cancer‐specific survival (CSS) and determination of structural break points with use of the Chow test (bottom row). The fitting bandwidth was 0.6. Each dot in the co‐plot represents an HR of a specific lymph node examined (LNE, vacant if case number < 10) from Cox regression analysis. ( a ) Overall patients, ( b ) adenocarcinoma patients, and ( c ) esophageal squamous cell carcinoma (ESCC) patients from the Surveillance, Epidemiology, and End Results (SEER) database. ( ) T1, ( ) T2, ( ) T3, and ( ) T4.

Article Snippet: **Curves were generated using odds ratios (ORs, each LNE count compared to 1 LNE as a reference) in logistic regression analysis and hazard ratios (HRs) in Cox regression analysis using Locally Weighted Scatterplot Smoothing (LOWESS, Stata 12.0) with a bandwidth of 0.6.

Techniques:

Journal: Thoracic Cancer

Article Title: Effect of lymph node examined count on accurate staging and survival of resected esophageal cancer

doi: 10.1111/1759-7714.13056

Figure Lengend Snippet: Co‐plot of hazard ratios (HRs, upper row) and Locally Weighted Scatterplot Smoothing curves of cancer‐specific survival (CSS) and determination of structural break points with the use of the Chow test (bottom row). The fitting bandwidth was 0.6. Each dot in the co‐plot represents an HR of a specific lymph node examined (LNE, vacant if case number < 10) from Cox regression analysis. ( a ) Node‐negative (N negative) and ( b ) node‐positive (N positive) patients. ( ) T1, ( ) T2, ( ) T3, and ( ) T4.

Article Snippet: **Curves were generated using odds ratios (ORs, each LNE count compared to 1 LNE as a reference) in logistic regression analysis and hazard ratios (HRs) in Cox regression analysis using Locally Weighted Scatterplot Smoothing (LOWESS, Stata 12.0) with a bandwidth of 0.6.

Techniques:

Journal: Science Advances

Article Title: Nitrate limitation in early Neoproterozoic oceans delayed the ecological rise of eukaryotes

doi: 10.1126/sciadv.ade9647

Figure Lengend Snippet: Red points in ( A ) to ( C ) denote samples with Fe T > 0.5%; blue points, Fe T 0.4 to 0.5%; black points, Fe T < 0.4%. Blue band in (B) marks sediments deposited under oxic conditions, and green band marks sediments deposited under anoxic conditions. In (C), green band represents ferruginous conditions, and orange band represents euxinic conditions. In ( D ), green band marks near-zero δ 15 N bulk values indicative of dominance of nitrogen fixation. Color and size of data points are keyed to reflect TN and TOC contents. Blue line is LOWESS regression fit with 95% CI (gray shadow) calculated from SEs.

Article Snippet: We note that more data from the 800- to 900-Ma time bin are needed to accurately define the rise of δ 15 N. Sensitivity tests of LOWESS regression and changepoint detection analysis on the lowest 50% data and the entire dataset (fig. S6) also revealed a stepwise rise at around 800 Ma, although these analyses show more temporal variation.

Techniques:

Journal: Science Advances

Article Title: Nitrate limitation in early Neoproterozoic oceans delayed the ecological rise of eukaryotes

doi: 10.1126/sciadv.ade9647

Figure Lengend Snippet: ( A ) Compilation of published nitrogen isotope data from Mesoproterozoic to Neoproterozoic. Boxplots show distribution of δ 15 N values in each 100-Ma time bin. Blue line is LOWESS regression fit of lowest 25% δ 15 N data in every 100-Ma time bin; see fig. S6 for LOWESS regression fit of lowest 50% data and the entire dataset. Gray shade is 95% CI calculated from 10,000 bootstrapping experiments. Red triangle marks stepwise increase from a changepoint analysis of lowest 25% δ 15 N data. Most significant change in both mean and variance occurs at ca. 820 Ma. The green shade marks δ 15 N values characteristic of a nitrogen cycle dominated by nitrogen fixation (e.g., f assimilator < 0.16; see Fig. 5 ). ( B ) Frequency distributions of δ 15 N values from bootstrapping experiments ( n = 10,000) of pre–800 Ma (blue) and post–800 Ma (orange) samples. ( C ) Frequency distributions of mean δ 15 N values from bootstrapping experiments ( n = 10,000) of pre–800 Ma (blue) and post–800 Ma (orange) samples.

Article Snippet: We note that more data from the 800- to 900-Ma time bin are needed to accurately define the rise of δ 15 N. Sensitivity tests of LOWESS regression and changepoint detection analysis on the lowest 50% data and the entire dataset (fig. S6) also revealed a stepwise rise at around 800 Ma, although these analyses show more temporal variation.

Techniques:

Journal: Science Advances

Article Title: Nitrate limitation in early Neoproterozoic oceans delayed the ecological rise of eukaryotes

doi: 10.1126/sciadv.ade9647

Figure Lengend Snippet: ( A ) LOWESS regression curve of lowest 25% δ 15 N data in every 100-Ma time bin. A stepwise rise in δ 15 N at ca. 800 Ma coincides with breakup of supercontinent Rodinia and implies increasing marine nitrate availability. ( B ) Phosphorus concentrations of marine siliciclastic sediments through time ( , ). LOWESS regression is based on the entire dataset. An increase in late Tonian implies greater phosphorus availability in the surface ocean. ( C ) Biomarker data (see the Supplementary Materials) . Note oldest steranes at ca. 820 Ma, increasing sterane/hopane ratios, and inferred dominance of various primary producers. ( D ) Approximate stratigraphic ranges of major eukaryotic clades .

Article Snippet: We note that more data from the 800- to 900-Ma time bin are needed to accurately define the rise of δ 15 N. Sensitivity tests of LOWESS regression and changepoint detection analysis on the lowest 50% data and the entire dataset (fig. S6) also revealed a stepwise rise at around 800 Ma, although these analyses show more temporal variation.

Techniques: Biomarker Discovery