Supplement 1. R source code for a two-source simulation mixing model that detects 15N-enriched seed in seed batches composed of unknown quantities of 15N-enriched (marked) and normal (unmarked) seeds.

File List Isotope_mix.R (md5: 1ad414fab0125136b63280ca539c8e3e) – R source code for two-source mixing model<br> datos.csv (md5: 9c6bd3134e2b86bb3b2db0b129ceb1a5) – data file containing of ¹⁵N atom % values and number of seeds per batch Description File contains the R source code Isotope_mix.R, a program to estimate the number of marked (¹⁵N- enriched) and unmarked (i.e., normal) seeds in a batch of n seeds. The program looks for the combination of marked and unmarked seeds that maximized the likelihood of observing a value of ¹⁵N atom % similar to the observed one. The empirical distributions of ¹⁵N from a particular combination of marked and unmarked seeds were estimated by simulating 10⁶ (a million) such seed mixtures by sampling from the combinations of Gaussian distribution fitted to the positive and negative controls (with truncation for the marked ones). Confidence intervals were found by looking at those combinations of marked and un-marked seeds that where 2 likelihood units from the maximum.

文件清单：
Isotope_mix.R（MD5: 1ad414fab0125136b63280ca539c8e3e）——用于双源混合模型的R语言源代码

datos.csv（MD5: 9c6bd3134e2b86bb3b2db0b129ceb1a5）——包含¹⁵N原子百分比数值与每批次种子数量的数据文件

数据集说明：
本数据集包含用于估算n粒种子批次中标记（¹⁵N富集）与未标记（即天然）种子数量的R语言源代码Isotope_mix.R。该程序通过寻找标记与未标记种子的最优组合，以最大化观测到与实测¹⁵N原子百分比值相近的似然性。针对特定标记与未标记种子组合的¹⁵N经验分布，通过以下方式估算：对适配于阳性与阴性对照的高斯分布（针对标记样本进行截尾处理）进行抽样，模拟10⁶（即一百万）组此类种子混合样本。置信区间则通过选取与最大似然值相差2个似然单位的标记与未标记种子组合来确定。