Weedy rice (Oryza sativa f. spontanea) is considered to be a pest in modern rice production systems because it competes for resources, has poor yield characteristics, and subsequently has a negative effect on rice grain yield. The evolutionary relationships among weedy rice, landrace rice, improved rice cultivars, and wild rice are largely unknown. In this study, we conducted a population genetic analysis based on neutral markers and gene haplotypes in 524 rice accessions and a comparative transcriptomic analysis using 15 representative samples. The results showed that weedy rice populations have the highest level of genetic diversity (He=0.8386), and can be divided into two groups (japonica-type and indica-type). The japonica-type weedy rice accessions from HLJ, JL, LN, and NX provinces clustered with the landraces grown in these same provinces. The indica-types from JS province also clustered with the indica-type landraces from JS province. Comparative transcriptome analysis of weedy rice populations, improved rice populations. and landrace rice from HLJ, JL and LN provinces showed that the weedy rice still clustered with the landrace rice, and that the improved rice lines comprise a single population. Thirty-two differentially expressed genes were shared by the improved rice and landrace rice groups as well as between the improved rice and weedy rice groups. Using GO analysis, we identified 19 shared GO terms in the improved rice and landrace rice groups as well as between the improved rice and weedy rice groups. Our results suggest that weedy rice populations in China have diverse origins, and comparative transcriptome analysis of different types of rice from HLJ, JL, and LN  suggests that improved rice populations have become a medium or end point in the evolution of weedy rice, which provides a new perspective for the study of weedy rice origins and lays a solid foundation for rice breeding.

The weedy rice samples were collected in improved rice fields, and then were planted and depuration in experiment fields in Beijing, The depuration samples have the stable phenotype to be used to proceed the study.

These data is the excel documents, and easy to understand.

The caption of supplemental materials:

SFig1 The distribution map of 524 samples
SFig2 The structure plot at the best K value of each population. 
Table S1 All primary data based on SSR markers.
Table S2a The calculation of the best K value for each population
Table S2 All SNPs from 15 samples.
Table S3 DEG analysis between weedy rice and landrace rice.
Table S4 DEG analysis between weedy rice and improved rice.
Table S5 DEG analysis between landrace rice and improved rice.
Table S6 DEG-Go term analysis between weedy rice and landrace rice. 
Table S7 DEG-Go term analysis between weedy rice and improved rice.
Table S8 DEG-Go term analysis between landrace rice and improved rice.