In this thesis we have studied the molecular characteristics of various types of non-Hodgkin lymphoma with the aim to advance our understanding of its biology and improve classification and disease diagnosis. For this, we have collected several lymphoma sample series on which we performed next generation sequencing (NGS). To interpret the large amount of NGS data we developed multiple bioinformatics pipelines, to study copy number aberration (CNA), mutation, translocation and immunoglobulin clonality analysis. Chapter 2 concerns breast implant-associated anaplastic large cell lymphoma (BIA-ALCL). BIA-ALCL is an ALK-negative anaplastic large T-cell lymphoma occurring in women with breast implants, which was at the time of the study a provisional entity in the WHO Classification. Our aim was to determine if specific molecular characteristics of BIA-ALCL could be identified, and to study its biological relation to other members of the ALCL family to ultimately answer the question whether BIA-ALCL would indeed be a separate entity. A characteristic loss of chromosome 20q13.13 was observed in 66% of BIA-ALCL samples versus 13% in ALK-negative ALCL. These observations justify recognition of BIAALCL as a separate disease entity, that should be discriminated from other ALCL entities. In chapter 3 we studied follicular lymphoma (FL). The clinical behavior within FL is diverse, stage I FL may be cured in ∼50% of cases, while stage III/IV is in general incurable. The questions we asked were: if this is because stage I FL is detected already in an early phase of the disease, or are there essential differences in the underlying biology? Chapter 4 is also dedicated to FL. In this chapter we compared two other ends of the clinical spectrums. Our aim was to predict which FL patients do require treatment immediate after primary diagnosis and for which patients treatment can be delayed for an extended period of time (years). Opposite to our expectation, there were no differences observed in frequencies of alterations. In chapter 5 diffuse large B-cell lymphoma (DLBCL) was studied. DLBCL is a heterogeneous disease. Our aim was to evaluate the predictive power of the international prognostic index (IPI), and interim and end-of-treatment (eot) 18F-FDG-positron emission tomography (PET)/ computer tomography (CT) in the various molecular DLBCL classes. A significantly worse survival was found for DFCIs’ cluster C5 patients and NCI’s Lymphgen MCD counterpart compared to other classes and independent of IPI. We found that despite a negative eot PET/CT after 6 to 8 cycles of treatment, hence initial complete metabolic response to standard immunochemotherapy, C5/MCD class of patients had a significantly worse prognosis, compared to all other molecular classes. In chapter 6 large B-cell lymphoma of immune privileged sites (LBCL-IP) were analyzed. We studied their molecular landscape and used this to derive the clonal relationships between primary and relapsed LBCL-IP. LBCL-IP is a B-cell lymphoma located at primary diagnosis in immune sanctuaries, including the testis and the central nervous system (CNS). These tumors have a poor prognosis with relapses in almost 50% of patients, even with a long interval after reaching complete response. All LBCL-IP pairs were shown to be clonally related. Nevertheless, on average only 24% of the mutations were shared, indicating parallel tumor evolution. Treatment regimens may be improved with this knowledge, aiming for treatment of the CPC and longtime follow up monitoring. In chapter 7 FFPE-targeted locus capture (TLC), is presented together with a bioinformatics implementation for subsequent rearrangement detection. This technique showed to be more sensitive and specific than fluorescent in situ hybridization (FISH), and a better performance was observed in highly repetitive regions compared to translocation detection with standard capture next generation sequencing (NGS).